Search

In this study, the Quantum-Train Quantum Fast Weight Programmer (QT-QFWP) framework is proposed, which facilitates the efficient and scalable programming of variational quantum circuits (VQCs) by leveraging quantum-driven parameter updates for the classical slow programmer that controls the fast programmer VQC model. This approach offers a significant advantage over conventional hybrid quantum-classical models by optimizing both quantum and classical parameter management. The framework has been benchmarked across several time-series prediction tasks, including Damped Simple Harmonic Motion (SHM), NARMA5, and Simulated Gravitational Waves (GW), demonstrating its ability to reduce parameters by roughly 70-90\% compared to Quantum Long Short-term Memory (QLSTM) and Quantum Fast Weight Programmer (QFWP) without compromising accuracy. The results show that QT-QFWP outperforms related models in both efficiency and predictive accuracy, providing a pathway toward more practical and cost-effective quantum machine learning applications. This innovation is particularly promising for near-term quantum systems, where limited qubit resources and gate fidelities pose significant constraints on model complexity. QT-QFWP enhances the feasibility of deploying VQCs in time-sensitive applications and broadens the scope of quantum computing in machine learning domains., Comment: 9 pages, 7 figures

Recent studies have successfully integrated large vision-language models (VLMs) into low-level robotic control by supervised fine-tuning (SFT) with expert robotic datasets, resulting in what we term vision-language-action (VLA) models. Although the VLA models are powerful, how to improve these large models during interaction with environments remains an open question. In this paper, we explore how to further improve these VLA models via Reinforcement Learning (RL), a commonly used fine-tuning technique for large models. However, we find that directly applying online RL to large VLA models presents significant challenges, including training instability that severely impacts the performance of large models, and computing burdens that exceed the capabilities of most local machines. To address these challenges, we propose iRe-VLA framework, which iterates between Reinforcement Learning and Supervised Learning to effectively improve VLA models, leveraging the exploratory benefits of RL while maintaining the stability of supervised learning. Experiments in two simulated benchmarks and a real-world manipulation suite validate the effectiveness of our method., Comment: Accepted to ICRA 2025

Human motion generation has advanced markedly with the advent of diffusion models. Most recent studies have concentrated on generating motion sequences based on text prompts, commonly referred to as text-to-motion generation. However, the bidirectional generation of motion and text, enabling tasks such as motion-to-text alongside text-to-motion, has been largely unexplored. This capability is essential for aligning diverse modalities and supports unconditional generation. In this paper, we introduce PackDiT, the first diffusion-based generative model capable of performing various tasks simultaneously, including motion generation, motion prediction, text generation, text-to-motion, motion-to-text, and joint motion-text generation. Our core innovation leverages mutual blocks to integrate multiple diffusion transformers (DiTs) across different modalities seamlessly. We train PackDiT on the HumanML3D dataset, achieving state-of-the-art text-to-motion performance with an FID score of 0.106, along with superior results in motion prediction and in-between tasks. Our experiments further demonstrate that diffusion models are effective for motion-to-text generation, achieving performance comparable to that of autoregressive models.

Large Language Models (LLMs) have demonstrated remarkable abilities in tackling a wide range of complex tasks. However, their huge computational and memory costs raise significant challenges in deploying these models on resource-constrained devices or efficiently serving them. Prior approaches have attempted to alleviate these problems by permanently removing less important model structures, yet these methods often result in substantial performance degradation due to the permanent deletion of model parameters. In this work, we tried to mitigate this issue by reducing the number of active parameters without permanently removing them. Specifically, we introduce a differentiable dynamic pruning method that pushes dense models to maintain a fixed number of active parameters by converting their MLP layers into a Mixture of Experts (MoE) architecture. Our method, even without fine-tuning, consistently outperforms previous structural pruning techniques across diverse model families, including Phi-2, LLaMA-2, LLaMA-3, and Qwen-2.5.

Quantum chemistry is among the most promising applications of quantum computing, offering the potential to solve complex electronic structure problems more efficiently than classical approaches. A critical component of any quantum algorithm is the measurement step, where the desired properties are extracted from a quantum computer. This review focuses on recent advancements in quantum measurement techniques tailored for quantum chemistry, particularly within the second quantized framework suitable for current and near-term quantum hardware. We provide a comprehensive overview of measurement strategies developed primarily for the Variational Quantum Eigensolver (VQE) and its derivatives. These strategies address the inherent challenges posed by complexity of the electronic Hamiltonian operator. Additionally, we examine methods for estimating excited states and one- and two-electron properties, extending the applicability of quantum algorithms to broader chemical phenomena. Key aspects of the review include approaches for constructing measurement operators with reduced classical preprocessing and quantum implementation costs, techniques to minimize the number of measurements required for a given accuracy, and error mitigation strategies that leverage symmetries and other properties of the measurement operators. Furthermore, we explore measurement schemes rooted in Quantum Phase Estimation (QPE), which are expected to become viable with the advent of fault-tolerant quantum computing. This review emphasizes foundational concepts and methodologies rather than numerical benchmarks, serving as a resource for researchers aiming to enhance the efficiency and accuracy of quantum measurements in quantum chemistry.

Maintaining consistent model performance across domains is a fundamental challenge in machine learning. While recent work has explored using LLM-generated data for fine-tuning, its impact on cross-domain generalization remains poorly understood. In this paper, we present a systematic analysis revealing that fine-tuning with LLM-generated data not only improves target task performance but also reduces out-of-domain (OOD) degradation compared to fine-tuning with ground truth data. Through analyzing the data sequence in tasks of various domains, we demonstrate that this enhanced OOD robustness stems from a reduced prevalence of high perplexity tokens in LLM-generated sequences. Following this hypothesis we showed that masking high perplexity tokens in ground truth training data also achieves similar OOD preservation comparable to using LLM-generated data. Extensive experiments across diverse model architectures and scales, including Gemma2-2B, Mistral-7B and Llama3-8B, corroborate the consistency of our findings. To the best of our knowledge, this work provides the first mechanistic explanation for the superior OOD robustness conferred by LLM-generated training data, offering valuable insights for developing more robust fine-tuning strategies.

The rapid proliferation of large language models (LLMs) in natural language processing (NLP) has created a critical need for techniques that enable efficient deployment on memory-constrained devices without compromising performance. We present a method to prune LLMs that selectively prunes model blocks based on an importance score and replaces them with a low-parameter replacement strategy. Specifically, we propose a principled metric to replace each pruned block using a weight-sharing mechanism that leverages unpruned counterparts from the model and block-specific low-rank adapters. Furthermore, we facilitate the learning of these replacement blocks with output feature normalization and an adapter initialization scheme built on low-rank SVD reconstructions. Empirical evaluations demonstrate substantial performance gains over existing methods, achieving state-of-the-art performance on 5/6 benchmarks for a compression rate of 30% and 6/6 benchmarks for a compression rate of 40%. We also demonstrate that our approach can extend smaller models, boosting performance on 6/6 benchmarks using only ~0.3% tokens of extended training with minimal additional parameter costs., Comment: Accepted to NAACL 2025 - Main Conference

The need to explore and/or optimize expensive simulators with many qualitative factors arises in broad scientific and engineering problems. Our motivating application lies in path planning - the exploration of feasible paths for navigation, which plays an important role in robotics, surgical planning and assembly planning. Here, the feasibility of a path is evaluated via expensive virtual experiments, and its parameter space is typically discrete and high-dimensional. A carefully selected experimental design is thus essential for timely decision-making. We propose here a novel framework, called QuIP, for experimental design of Qualitative factors via Integer Programming under a Gaussian process surrogate model with an exchangeable covariance function. For initial design, we show that its asymptotic D-optimal design can be formulated as a variant of the well-known assignment problem in operations research, which can be efficiently solved to global optimality using state-of-the-art integer programming solvers. For sequential design (specifically, for active learning or black-box optimization), we show that its design criterion can similarly be formulated as an assignment problem, thus enabling efficient and reliable optimization with existing solvers. We then demonstrate the effectiveness of QuIP over existing methods in a suite of path planning experiments and an application to rover trajectory optimization.

Qualitative structure-activity relationship (QSAR) is important for drug discovery and offers valuable insights into the biological interactions of potential drug candidates. It has been demonstrated that QSAR can be accurately predicted by machine learning. However, data with poor quality and limited availability are always the most common and critical issues for medical-related applications for machine learning. In this manuscript, we aim to discuss the performance of classical and quantum classifiers in QSAR prediction and attempt to demonstrate the quantum advantages in the generalization power of the quantum classifier under conditions of limited data availability and a reduced number of features. By applying different data embedding methods followed by feature selection through principal component analysis (PCA), we find that the quantum classifier outperforms the classical one when a small number of features are selected and the number of training samples is limited. The generality of quantum advantages in other open datasets is also explored.

In the realm of online advertising, optimizing conversions is crucial for delivering relevant products to users and enhancing business outcomes. Predicting conversion events is challenging due to variable delays between user interactions, such as impressions or clicks, and the actual conversions. These delays differ significantly across various advertisers and products, necessitating distinct optimization time windows for targeted conversions. To address this, we introduce a novel approach named the \textit{Personalized Interpolation} method, which innovatively builds upon existing fixed conversion window models to estimate flexible conversion windows. This method allows for the accurate estimation of conversions across a variety of delay ranges, thus meeting the diverse needs of advertisers without increasing system complexity. To validate the efficacy of our proposed method, we conducted comprehensive experiments using ads conversion model. Our experiments demonstrate that this method not only achieves high prediction accuracy but also does so more efficiently than other existing solutions. This validation underscores the potential of our Personalized Interpolation method to significantly enhance conversion optimization in real-world online advertising systems, promising improved targeting and effectiveness in advertising strategies.

Image restoration aims to recover content from inputs degraded by various factors, such as adverse weather, blur, and noise. Perceptual Image Restoration (PIR) methods improve visual quality but often do not support downstream tasks effectively. On the other hand, Task-oriented Image Restoration (TIR) methods focus on enhancing image utility for high-level vision tasks, sometimes compromising visual quality. This paper introduces UniRestore, a unified image restoration model that bridges the gap between PIR and TIR by using a diffusion prior. The diffusion prior is designed to generate images that align with human visual quality preferences, but these images are often unsuitable for TIR scenarios. To solve this limitation, UniRestore utilizes encoder features from an autoencoder to adapt the diffusion prior to specific tasks. We propose a Complementary Feature Restoration Module (CFRM) to reconstruct degraded encoder features and a Task Feature Adapter (TFA) module to facilitate adaptive feature fusion in the decoder. This design allows UniRestore to optimize images for both human perception and downstream task requirements, addressing discrepancies between visual quality and functional needs. Integrating these modules also enhances UniRestore's adapability and efficiency across diverse tasks. Extensive expertments demonstrate the superior performance of UniRestore in both PIR and TIR scenarios., Comment: 11 pages, 6 figures

Our research aims to develop machines that learn to perceive visual motion as do humans. While recent advances in computer vision (CV) have enabled DNN-based models to accurately estimate optical flow in naturalistic images, a significant disparity remains between CV models and the biological visual system in both architecture and behavior. This disparity includes humans' ability to perceive the motion of higher-order image features (second-order motion), which many CV models fail to capture because of their reliance on the intensity conservation law. Our model architecture mimics the cortical V1-MT motion processing pathway, utilizing a trainable motion energy sensor bank and a recurrent graph network. Supervised learning employing diverse naturalistic videos allows the model to replicate psychophysical and physiological findings about first-order (luminance-based) motion perception. For second-order motion, inspired by neuroscientific findings, the model includes an additional sensing pathway with nonlinear preprocessing before motion energy sensing, implemented using a simple multilayer 3D CNN block. When exploring how the brain acquired the ability to perceive second-order motion in natural environments, in which pure second-order signals are rare, we hypothesized that second-order mechanisms were critical when estimating robust object motion amidst optical fluctuations, such as highlights on glossy surfaces. We trained our dual-pathway model on novel motion datasets with varying material properties of moving objects. We found that training to estimate object motion from non-Lambertian materials naturally endowed the model with the capacity to perceive second-order motion, as can humans. The resulting model effectively aligns with biological systems while generalizing to both first- and second-order motion phenomena in natural scenes.

This paper presents a comprehensive analysis of power plant performance using the inverse Gaussian (IG) distribution framework. We combine theoretical foundations with practical applications, focusing on both combined cycle and nuclear power plant contexts. The study demonstrates the advantages of the IG distribution in modeling right-skewed industrial data, particularly in power generation. Using the UCI Combined Cycle Power Plant Dataset, we establishthe superiority of IG-based models over traditional approaches through rigorous statistical testing and model validation. The methodology developed here extends naturally to nuclear power plant applications, where similar statistical patterns emerge in operational data. Our findings suggest that IG-based models provide more accurate predictions and better capture the underlying physical processes in power generation systems.

With the rapid growth of generative AI in numerous applications, explainable AI (XAI) plays a crucial role in ensuring the responsible development and deployment of generative AI technologies. XAI has undergone notable advancements and widespread adoption in recent years, reflecting a concerted push to enhance the transparency, interpretability, and credibility of AI systems. Recent research emphasizes that a proficient XAI method should adhere to a set of criteria, primarily focusing on two key areas. Firstly, it should ensure the quality and fluidity of explanations, encompassing aspects like faithfulness, plausibility, completeness, and tailoring to individual needs. Secondly, the design principle of the XAI system or mechanism should cover the following factors such as reliability, resilience, the verifiability of its outputs, and the transparency of its algorithm. However, research in XAI for generative models remains relatively scarce, with little exploration into how such methods can effectively meet these criteria in that domain. In this work, we propose PXGen, a post-hoc explainable method for generative models. Given a model that needs to be explained, PXGen prepares two materials for the explanation, the Anchor set and intrinsic & extrinsic criteria. Those materials are customizable by users according to their purpose and requirements. Via the calculation of each criterion, each anchor has a set of feature values and PXGen provides examplebased explanation methods according to the feature values among all the anchors and illustrated and visualized to the users via tractable algorithms such as k-dispersion or k-center.

Large language models (LLMs) have recently demonstrated remarkable success in mathematical reasoning. Despite progress in methods like chain-of-thought prompting and self-consistency sampling, these advances often focus on final correctness without ensuring that the underlying reasoning process is coherent and reliable. This paper introduces Step-KTO, a training framework that combines process-level and outcome-level binary feedback to guide LLMs toward more trustworthy reasoning trajectories. By providing binary evaluations for both the intermediate reasoning steps and the final answer, Step-KTO encourages the model to adhere to logical progressions rather than relying on superficial shortcuts. Our experiments on challenging mathematical benchmarks show that Step-KTO significantly improves both final answer accuracy and the quality of intermediate reasoning steps. For example, on the MATH-500 dataset, Step-KTO achieves a notable improvement in Pass@1 accuracy over strong baselines. These results highlight the promise of integrating stepwise process feedback into LLM training, paving the way toward more interpretable and dependable reasoning capabilities.

Accurate network data are essential in fields such as economics, sociology, and computer science. Aggregated Relational Data (ARD) provides a way to capture network structures using partial data. This article compares two main frameworks for recovering network links from ARD: Bayesian Latent Surface Modeling (BLSM) and Frequentist Penalized Regression (FPR). Using simulation studies and real-world applications, we evaluate their theoretical properties, computational efficiency, and practical utility in domains like financial risk assessment and epidemiology. Key findings emphasize the importance of trait design, privacy considerations, and hybrid modeling approaches to improve scalability and robustness.

We present a new photometric pipeline for the detection of pre-supernova (pre-SN) emission in the Young Supernova Experiment (YSE) sky survey. The method described is applied to SN 2020tlf, a type II SN (SN II) with precursor emission in the last ~100 days before first light. We re-analyze the YSE griz-band light curves of SN 2020tlf and provide revised pre-explosion photometry that includes a robust list of confident detection and limiting magnitudes. Compared to the results of Jacobson-Galan et al. 2022a, this new analysis yields fewer total r/i/z-band pre-SN detections at phases > -100 days. Furthermore, we discourage the use of the blackbody modeling of the pre-explosion spectral energy distribution, the pre-SN bolometric light curve and the blackbody model parameters presented in Jacobson-Galan et al. 2022a. Nevertheless, binned photometry of SN 2020tlf confirms a consistent progenitor luminosity of ~10$^{40}$ erg s$^{-1}$ before explosion., Comment: 5 pages, 1 figure. Published in RNAAS

Axions are hypothetical pseudoscalar particles that have been regarded as promising dark matter (DM) candidates. On the other hand, extended compact objects such as axion stars, which are supported by gravity and axion self interactions, may have also been formed in the early Universe and comprise part of DM. In this work, we consider the lensing of electromagnetic signals from distant sources by axion stars, as a way to constrain the properties of axion stars and fundamental axion parameters. Accounting for the effect of the finite size of the axion star, we study the lensing effect induced by gravity and the axion-photon coupling. The latter effect is frequency dependent, and is relevant in the low frequency band, which motivates the use of fast radio burst (FRB) signals as a probe. We calculate the predicted number of lensed FRB events by specifying the fundamental axion parameters, axion star radial profile, fraction of DM residing in axion stars, and imposing lensing criteria based on the flux ratio and time delay between the brightest images from lensing. Assuming an optimistic case of $10^4$ observed FRB events, and a timing resolution of $1~\mu{\rm s}$, the lack of observed FRB lensing events in CHIME allows us to probe axion stars with mass $ \gtrsim 2 \times 10^{-2} M_\odot$, corresponding to axion masses $\lesssim 10^{-10}{\rm eV}$. We obtain constraints for even lighter axion stars up to $\sim 10^{-3} M_\odot$, when the axion-photon interactions are taken into account. Our results indicate that FRB lensing lead to constraints that are competitive with conventional microlensing searches operating in the optical band., Comment: 36 pages, 37 figures

Statistical methods for causal inference with continuous treatments mainly focus on estimating the mean potential outcome function, commonly known as the dose-response curve. However, it is often not the dose-response curve but its derivative function that signals the treatment effect. In this paper, we investigate nonparametric inference on the derivative of the dose-response curve with and without the positivity condition. Under the positivity and other regularity conditions, we propose a doubly robust (DR) inference method for estimating the derivative of the dose-response curve using kernel smoothing. When the positivity condition is violated, we demonstrate the inconsistency of conventional inverse probability weighting (IPW) and DR estimators, and introduce novel bias-corrected IPW and DR estimators. In all settings, our DR estimator achieves asymptotic normality at the standard nonparametric rate of convergence. Additionally, our approach reveals an interesting connection to nonparametric support and level set estimation problems. Finally, we demonstrate the applicability of our proposed estimators through simulations and a case study of evaluating a job training program., Comment: 111 pages (24 pages for the main paper), 9 figures

Long-range magnetic ordering of the orbital motion of oxygen atoms within NO$_3$$^-$ cations is identified from experimental measurements of the magnetic susceptibility $\chi$($T$) in LiNO$_3$, Ca(NO$_3$)$_2$, K$_{0.5}$Rb$_{0.5}$NO$_3$ and C(NH$_2$)$_3$NO$_3$ at their respective order-disorder, solid-solid phase transitions $T$$_N$. The observed sharp changes in $\chi$($T$) and accompanying hysteretic behavior indicate the phase transitions to be first order. A model employing the law of conservation of angular momentum is used to explain why the librations between neighboring NO$_3$$^-$ become geared below $T$$_N$. Since the periodic motions involve concerted motion of net charges, the associated magnetic moments of the NO$_3$$^-$ ions indirectly establish an antiferromagnetic structure below $T$$_N$. Our findings identify a previously unidentified type of molecular interaction which may be exploited to further increase the enthalpy of the widely-popular hydrated salts employed as energy storage devices., Comment: 13 pages (single column, 1.5 spaced), 5 figures

The rapid progress in quantum computing (QC) and machine learning (ML) has attracted growing attention, prompting extensive research into quantum machine learning (QML) algorithms to solve diverse and complex problems. Designing high-performance QML models demands expert-level proficiency, which remains a significant obstacle to the broader adoption of QML. A few major hurdles include crafting effective data encoding techniques and parameterized quantum circuits, both of which are crucial to the performance of QML models. Additionally, the measurement phase is frequently overlooked-most current QML models rely on pre-defined measurement protocols that often fail to account for the specific problem being addressed. We introduce a novel approach that makes the observable of the quantum system-specifically, the Hermitian matrix-learnable. Our method features an end-to-end differentiable learning framework, where the parameterized observable is trained alongside the ordinary quantum circuit parameters simultaneously. Using numerical simulations, we show that the proposed method can identify observables for variational quantum circuits that lead to improved outcomes, such as higher classification accuracy, thereby boosting the overall performance of QML models., Comment: Accepted by ICASSP 2025 Workshop: Quantum Machine Learning in Signal Processing and Artificial Intelligence

Existing benchmarks for evaluating long-context language models (LCLMs) primarily focus on long-context recall, requiring models to produce short responses based on a few critical snippets while processing thousands of irrelevant tokens. We introduce LongProc (Long Procedural Generation), a new benchmark that requires both the integration of highly dispersed information and long-form generation. LongProc consists of six diverse procedural generation tasks, such as extracting structured information from HTML pages into a TSV format and executing complex search procedures to create travel plans. These tasks challenge LCLMs by testing their ability to follow detailed procedural instructions, synthesize and reason over dispersed information, and generate structured, long-form outputs (up to 8K tokens). Furthermore, as these tasks adhere to deterministic procedures and yield structured outputs, they enable reliable rule-based evaluation. We evaluate 17 LCLMs on LongProc across three difficulty levels, with maximum numbers of output tokens set at 500, 2K, and 8K. Notably, while all tested models claim a context window size above 32K tokens, open-weight models typically falter on 2K-token tasks, and closed-source models like GPT-4o show significant degradation on 8K-token tasks. Further analysis reveals that LCLMs struggle to maintain long-range coherence in long-form generations. These findings highlight critical limitations in current LCLMs and suggest substantial room for improvement. Data and code available at: https://princeton-pli.github.io/LongProc

Physical AI needs to be trained digitally first. It needs a digital twin of itself, the policy model, and a digital twin of the world, the world model. In this paper, we present the Cosmos World Foundation Model Platform to help developers build customized world models for their Physical AI setups. We position a world foundation model as a general-purpose world model that can be fine-tuned into customized world models for downstream applications. Our platform covers a video curation pipeline, pre-trained world foundation models, examples of post-training of pre-trained world foundation models, and video tokenizers. To help Physical AI builders solve the most critical problems of our society, we make our platform open-source and our models open-weight with permissive licenses available via https://github.com/NVIDIA/Cosmos.

Reorienting polyatomic ions such as NH4+ and NO3- exhibit weak magnetic fields because the ions at the extremities trace out current loops; if the periodic reorientations become long-range ordered (i.e. gearing of neighboring NO3-), then the magnetic susceptibility should exhibit a unique signature along the different crystallographic axes. For the case of ammonium nitrate NH4NO3, we report the presence of two successive sharp steps in the molar magnetic susceptibility along the a- and b-axes upon crossing its order-disorder phase transition (from phase IV to phase II). We suggest the first step pertains to the NO3- planes shifting away from facing only along the b-axis and onto the a-axis by 45{\deg}. The second step is attributed to the disordering (ungearing) of the NH4+ and NO3-. In contrast, only one step was observed in the magnetic susceptibility along the c-axis and its large magnitude suggest the NO3- remain weakly correlated even in phase I at 400 K. We also find evidence that the NH4+ become magnetically ordered (geared) along the c-axis only until phase V. The approach employed in this work can be extended to experimentally study the lattice dynamics of other solids possessing planar ions such as amphidynamic crystals., Comment: 13 pages (single column), 4 figures

Vocabulary acquisition is essential to second language learning, as it underpins all core language skills. Accurate vocabulary assessment is particularly important in standardized exams, where test items evaluate learners' comprehension and contextual use of words. Previous research has explored methods for generating distractors to aid in the design of English vocabulary tests. However, current approaches often rely on lexical databases or predefined rules, and frequently produce distractors that risk invalidating the question by introducing multiple correct options. In this study, we focus on English vocabulary questions from Taiwan's university entrance exams. We analyze student response distributions to gain insights into the characteristics of these test items and provide a reference for future research. Additionally, we identify key limitations in how large language models (LLMs) support teachers in generating distractors for vocabulary test design. To address these challenges, we propose the iterative selection with self-review (ISSR) framework, which makes use of a novel LLM-based self-review mechanism to ensure that the distractors remain valid while offering diverse options. Experimental results show that ISSR achieves promising performance in generating plausible distractors, and the self-review mechanism effectively filters out distractors that could invalidate the question.

In the task of reference-based image inpainting, an additional reference image is provided to restore a damaged target image to its original state. The advancement of diffusion models, particularly Stable Diffusion, allows for simple formulations in this task. However, existing diffusion-based methods often lack explicit constraints on the correlation between the reference and damaged images, resulting in lower faithfulness to the reference images in the inpainting results. In this work, we propose CorrFill, a training-free module designed to enhance the awareness of geometric correlations between the reference and target images. This enhancement is achieved by guiding the inpainting process with correspondence constraints estimated during inpainting, utilizing attention masking in self-attention layers and an objective function to update the input tensor according to the constraints. Experimental results demonstrate that CorrFill significantly enhances the performance of multiple baseline diffusion-based methods, including state-of-the-art approaches, by emphasizing faithfulness to the reference images., Comment: WACV 2025. Project page: https://corrfill.github.io/

The deployment of Large Language Models (LLMs) in recommender systems for predicting Click-Through Rates (CTR) necessitates a delicate balance between computational efficiency and predictive accuracy. This paper presents an optimization framework that combines Retrieval-Augmented Generation (RAG) with an innovative multi-head early exit architecture to concurrently enhance both aspects. By integrating Graph Convolutional Networks (GCNs) as efficient retrieval mechanisms, we are able to significantly reduce data retrieval times while maintaining high model performance. The early exit strategy employed allows for dynamic termination of model inference, utilizing real-time predictive confidence assessments across multiple heads. This not only quickens the responsiveness of LLMs but also upholds or improves their accuracy, making it ideal for real-time application scenarios. Our experiments demonstrate how this architecture effectively decreases computation time without sacrificing the accuracy needed for reliable recommendation delivery, establishing a new standard for efficient, real-time LLM deployment in commercial systems.

This work analyzes transfer learning of the Variational Quantum Circuit (VQC). Our framework begins with a pretrained VQC configured in one domain and calculates the transition of 1-parameter unitary subgroups required for a new domain. A formalism is established to investigate the adaptability and capability of a VQC under the analysis of loss bounds. Our theory observes knowledge transfer in VQCs and provides a heuristic interpretation for the mechanism. An analytical fine-tuning method is derived to attain the optimal transition for adaptations of similar domains., Comment: Submitted to ICASSP 2025

We introduce intra-class memorability, where certain images within the same class are more memorable than others despite shared category characteristics. To investigate what features make one object instance more memorable than others, we design and conduct human behavior experiments, where participants are shown a series of images one at a time, and they must identify when the current item matches the item presented a few steps back in the sequence. To quantify memorability, we propose the Intra-Class Memorability score (ICMscore), a novel metric that incorporates the temporal intervals between repeated image presentations into its calculation. Our contributions open new pathways in understanding intra-class memorability by scrutinizing fine-grained visual features that result in the least and most memorable images and laying the groundwork for real-world applications in cognitive science and computer vision.

End-to-end encryption (E2EE) has become the gold standard for securing communications, bringing strong confidentiality and privacy guarantees to billions of users worldwide. However, the current push towards widespread integration of artificial intelligence (AI) models, including in E2EE systems, raises some serious security concerns. This work performs a critical examination of the (in)compatibility of AI models and E2EE applications. We explore this on two fronts: (1) the integration of AI "assistants" within E2EE applications, and (2) the use of E2EE data for training AI models. We analyze the potential security implications of each, and identify conflicts with the security guarantees of E2EE. Then, we analyze legal implications of integrating AI models in E2EE applications, given how AI integration can undermine the confidentiality that E2EE promises. Finally, we offer a list of detailed recommendations based on our technical and legal analyses, including: technical design choices that must be prioritized to uphold E2EE security; how service providers must accurately represent E2EE security; and best practices for the default behavior of AI features and for requesting user consent. We hope this paper catalyzes an informed conversation on the tensions that arise between the brisk deployment of AI and the security offered by E2EE, and guides the responsible development of new AI features.

This paper aims to achieve single-channel target speech extraction (TSE) in enclosures by solely utilizing distance information. This is the first work that utilizes only distance cues without using speaker physiological information for single-channel TSE. Inspired by recent single-channel Distance-based separation and extraction methods, we introduce a novel model that efficiently fuses distance information with time-frequency (TF) bins for TSE. Experimental results in both single-room and multi-room scenarios demonstrate the feasibility and effectiveness of our approach. This method can also be employed to estimate the distances of different speakers in mixed speech. Online demos are available at https://runwushi.github.io/distance-demo-page., Comment: 5 pages, 3 figures, accepted by ICASSP 2025

One of the common challenges faced by researchers in recent data analysis is missing values. In the context of penalized linear regression, which has been extensively explored over several decades, missing values introduce bias and yield a non-positive definite covariance matrix of the covariates, rendering the least square loss function non-convex. In this paper, we propose a novel procedure called the linear shrinkage positive definite (LPD) modification to address this issue. The LPD modification aims to modify the covariance matrix of the covariates in order to ensure consistency and positive definiteness. Employing the new covariance estimator, we are able to transform the penalized regression problem into a convex one, thereby facilitating the identification of sparse solutions. Notably, the LPD modification is computationally efficient and can be expressed analytically. In the presence of missing values, we establish the selection consistency and prove the convergence rate of the $\ell_1$-penalized regression estimator with LPD, showing an $\ell_2$-error convergence rate of square-root of $\log p$ over $n$ by a factor of $(s_0)^{3/2}$ ($s_0$: the number of non-zero coefficients). To further evaluate the effectiveness of our approach, we analyze real data from the Genomics of Drug Sensitivity in Cancer (GDSC) dataset. This dataset provides incomplete measurements of drug sensitivities of cell lines and their protein expressions. We conduct a series of penalized linear regression models with each sensitivity value serving as a response variable and protein expressions as explanatory variables.

As part of the BISTRO survey, we present JCMT 850 $\mu$m polarimetric observations towards the Orion Integral-Shaped Filament (ISF) that covers three portions known as OMC-1, OMC-2, and OMC-3. The magnetic field threading the ISF seen in the JCMT POL-2 map appears as a tale of three: pinched for OMC-1, twisted for OMC-2, and nearly uniform for OMC-3. A multi-scale analysis shows that the magnetic field structure in OMC-3 is very consistent at all the scales, whereas the field structure in OMC-2 shows no correlation across different scales. In OMC-1, the field retains its mean orientation from large to small scales, but shows some deviations at small scales. Histograms of relative orientations between the magnetic field and filaments reveal a bimodal distribution for OMC-1, a relatively random distribution for OMC-2, and a distribution with a predominant peak at 90$^\circ$ for OMC-3. Furthermore, the magnetic fields in OMC-1 and OMC-3 both appear to be aligned perpendicular to the fibers, which are denser structures within the filament, but the field in OMC-2 is aligned along with the fibers. All these suggest that gravity, turbulence, and magnetic field are each playing a leading role in OMC-1, 2, and 3, respectively. While OMC-2 and 3 have almost the same gas mass, density, and non-thermal velocity dispersion, there are on average younger and fewer young stellar objects in OMC-3, providing evidence that a stronger magnetic field will induce slower and less efficient star formation in molecular clouds., Comment: published in the ApJ Letters

Recently, energy-energy correlators (EECs) have garnered renewed interest for studying hadronic collisions at the Large Hadron Collider (LHC) and the Relativistic Heavy Ion Collider (RHIC). EEC measurements within jets provide a clear scale separation, facilitating the study of both perturbative and non-perturbative Quantum Chromodynamics (QCD) in the collinear limit. These proceedings present recent EEC results from the archived ALEPH $e^{+}e^{-}$ data taken at LEP at $\sqrt{s}$ = 91.2 GeV. In $e^{+}e^{-}$ collisions, perturbative and non-perturbative QCD can be studied with EECs in both the collinear limit using jets and the back-to-back limit using all particles as well as the transition between these two regimes. Comparisons of these results to generators and future extensions of this work will also be discussed., Comment: 4 pages, 1 figure, Proceedings of the 12th Large Hadron Collider Physics Conference

Large Vision Language Models (LVLMs) have achieved significant progress in integrating visual and textual inputs for multimodal reasoning. However, a recurring challenge is ensuring these models utilize visual information as effectively as linguistic content when both modalities are necessary to formulate an accurate answer. We hypothesize that hallucinations arise due to the lack of effective visual grounding in current LVLMs. This issue extends to vision-language benchmarks, where it is difficult to make the image indispensable for accurate answer generation, particularly in vision question-answering tasks. In this work, we introduce FiVL, a novel method for constructing datasets designed to train LVLMs for enhanced visual grounding and to evaluate their effectiveness in achieving it. These datasets can be utilized for both training and assessing an LVLM's ability to use image content as substantive evidence rather than relying solely on linguistic priors, providing insights into the model's reliance on visual information. To demonstrate the utility of our dataset, we introduce an innovative training task that outperforms baselines alongside a validation method and application for explainability. The code is available at https://github.com/IntelLabs/fivl.

Recent advancements in robotics have focused on developing generalist policies capable of performing multiple tasks. Typically, these policies utilize pre-trained vision encoders to capture crucial information from current observations. However, previous vision encoders, which trained on two-image contrastive learning or single-image reconstruction, can not perfectly capture the sequential information essential for embodied tasks. Recently, video diffusion models (VDMs) have demonstrated the capability to accurately predict future image sequences, exhibiting a good understanding of physical dynamics. Motivated by the strong visual prediction capabilities of VDMs, we hypothesize that they inherently possess visual representations that reflect the evolution of the physical world, which we term predictive visual representations. Building on this hypothesis, we propose the Video Prediction Policy (VPP), a generalist robotic policy conditioned on the predictive visual representations from VDMs. To further enhance these representations, we incorporate diverse human or robotic manipulation datasets, employing unified video-generation training objectives. VPP consistently outperforms existing methods across two simulated and two real-world benchmarks. Notably, it achieves a 28.1\% relative improvement in the Calvin ABC-D benchmark compared to the previous state-of-the-art and delivers a 28.8\% increase in success rates for complex real-world dexterous manipulation tasks., Comment: The first two authors contribute equally. Project Page at https://video-prediction-policy.github.io/

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

Recent advancements in quantum computing (QC) and machine learning (ML) have garnered significant attention, leading to substantial efforts toward the development of quantum machine learning (QML) algorithms to address a variety of complex challenges. The design of high-performance QML models, however, requires expert-level knowledge, posing a significant barrier to the widespread adoption of QML. Key challenges include the design of data encoding mechanisms and parameterized quantum circuits, both of which critically impact the generalization capabilities of QML models. We propose a novel method that encodes quantum circuit architecture information to enable the evolution of quantum circuit designs. In this approach, the fitness function is based on the effective dimension, allowing for the optimization of quantum circuits towards higher model capacity. Through numerical simulations, we demonstrate that the proposed method is capable of discovering variational quantum circuit architectures that offer improved learning capabilities, thereby enhancing the overall performance of QML models for complex tasks., Comment: Accepted by IEEE Symposium Series on Computational Intelligence - IEEE SSCI 2025

Neutrinos carry most of the energy released by a core-collapse supernova. SNOLAB has two neutrino-capable detectors, SNO+ and HALO, that have complementary neutrino flavour sensitivities. SNOLAB is also host to existing facilities, or plans to host future projects, that can enhance sensitivity to these neutrinos. These detectors, together with others worldwide both in existence and planned, will provide insights to a variety of different models using neutrinos from the next galactic supernova., Comment: 13 pages, 5 figures, for Astroparticle Physics special issue of Canadian Journal of Physics

The recent results of An, Luan, and Yen [Differential stability in convex optimization via generalized polyhedrality. Vietnam J. Math. https://-doi.org/10.1007/s10013-024-00721-y] on differential stability of parametric optimization problems described by proper generalized polyhedral convex functions and generalized polyhedral convex set-valued maps are analyzed, developed, and sharpened in this paper. Namely, keeping the Hausdorff locally convex topological vector spaces setting, we clarify the relationships between the upper estimates and lower estimates for the subdifferential and the singular subdifferential of the optimal value function. As shown by an example, the lower estimates can be strict. But, surprisingly, each upper estimate is an equality. Thus, exact formulas for the subdifferential and the singular subdifferential under consideration are obtained. In addition, it is proved that each subdifferential upper estimate coincides with the corresponding lower estimate if either the objective function or the constraint set-valued map is polyhedral convex.

We explored in this work the ubiquitous phenomenon of serial scammers, who deploy thousands of addresses to conduct a series of similar Rug Pulls on popular decentralized exchanges (DEXs). We first constructed a list of about 384,000 scammer addresses behind all 1-day Rug Pulls on the two most popular DEXs, Uniswap (Ethereum) and Pancakeswap (BSC), and identified many distinctive scam patterns including star-shaped, chain-shaped, and majority-flow scam clusters. We then proposed an algorithm to build a complete scam network from given scammer addresses, which consists of not only scammer addresses but also supporting addresses including depositors, withdrawers, transferrers, coordinators, and most importantly, wash traders. We note that profit estimations in existing works on Rug Pulls failed to capture the cost of wash trading, leading to inflated figures. Knowing who the wash traders are, we established a more accurate estimate for the true profit of individual scam pools as well as of the entire (serial) scam network by taking into account the wash-trading expenses.

Kinship face synthesis is a challenging problem due to the scarcity and low quality of the available kinship data. Existing methods often struggle to generate descendants with both high diversity and fidelity while precisely controlling facial attributes such as age and gender. To address these issues, we propose the Style Latent Diffusion Transformer (StyleDiT), a novel framework that integrates the strengths of StyleGAN with the diffusion model to generate high-quality and diverse kinship faces. In this framework, the rich facial priors of StyleGAN enable fine-grained attribute control, while our conditional diffusion model is used to sample a StyleGAN latent aligned with the kinship relationship of conditioning images by utilizing the advantage of modeling complex kinship relationship distribution. StyleGAN then handles latent decoding for final face generation. Additionally, we introduce the Relational Trait Guidance (RTG) mechanism, enabling independent control of influencing conditions, such as each parent's facial image. RTG also enables a fine-grained adjustment between the diversity and fidelity in synthesized faces. Furthermore, we extend the application to an unexplored domain: predicting a partner's facial images using a child's image and one parent's image within the same framework. Extensive experiments demonstrate that our StyleDiT outperforms existing methods by striking an excellent balance between generating diverse and high-fidelity kinship faces.

The rapid advancement of Vision-Language Models (VLMs) has significantly advanced the development of Embodied Question Answering (EQA), enhancing agents' abilities in language understanding and reasoning within complex and realistic scenarios. However, EQA in real-world scenarios remains challenging, as human-posed questions often contain noise that can interfere with an agent's exploration and response, bringing challenges especially for language beginners and non-expert users. To address this, we introduce a NoisyEQA benchmark designed to evaluate an agent's ability to recognize and correct noisy questions. This benchmark introduces four common types of noise found in real-world applications: Latent Hallucination Noise, Memory Noise, Perception Noise, and Semantic Noise generated through an automated dataset creation framework. Additionally, we also propose a 'Self-Correction' prompting mechanism and a new evaluation metric to enhance and measure both noise detection capability and answer quality. Our comprehensive evaluation reveals that current EQA agents often struggle to detect noise in questions, leading to responses that frequently contain erroneous information. Through our Self-Correct Prompting mechanism, we can effectively improve the accuracy of agent answers.

Since its inception in the 1960s, the internet has profoundly transformed human life. However, its original design now struggles to meet the evolving demands of modern society. Three primary defects have emerged: First, the concentration of power among a few dominant entities has intensified international conflicts and widened the technological divide. Second, the Internet Protocol (IP)-based system lacks inherent security, leading to frequent global cybersecurity incidents. Third, the rigidity of the IP protocol has hindered the sustainable development of cyberspace, as it resists necessary adaptations and innovations. Addressing these issues is crucial for the future resilience and security of the global digital landscape. To address these challenges, we propose the Co-governed Multi-Identifier Network (CoG-MIN briefly as MIN), a novel network architecture that leverages blockchain technology to ensure equal participation of countries worldwide in cyberspace governance and the rule of law. As a next-generation network system, CoG-MIN integrates mechanisms such as user authentication, data signatures, and encryption to significantly enhance network security. In testing environments, CoG-MIN has consistently withstood extensive attacks during various international cybersecurity competitions. Additionally, CoG-MIN supports the evolution and interoperability of different identifier systems, remains IP-compatible, and facilitates a gradual transition away from IP, providing an adaptable ecosystem for diverse network architectures. This adaptability fosters the development and evolution of diverse network architectures within CoG-MIN, making it a natural progression for the internet's future development. We further introduce a trilogy of cyberspace security theorems... (Due to character limitations, the full abstract is available in the paper PDF.)

For many pulsars, the scattering structures responsible for scintillation are typically dominated by a single, thin screen along the line of sight, which persists for years or decades. In recent years, an increasing number of doubly-lensed events have been observed, where a secondary lens crosses the line of sight. This causes additional or distorted scintillation arcs over time scales ranging from days to months. In this work we report such a transient event for pulsar B1737+13 and propose a possible lensing geometry including the distance to both lenses, and the orientation of the main screen. Using phase retrieval techniques to separate the two lenses in the wavefield, we report a curvature and rate of motion of features associated with the secondary lens as it passed through the line of sight. By fitting the annual variation of the curvature, we report a possible distance and orientation for the main screen. The distance of the secondary lens is found by mapping the secondary feature onto the sky and tracking its position over time for different distances. We validate this method using B0834+06, for which the screen solutions are known through VLBI, and successfully recover the correct solution for the secondary feature. With the identified lensing geometry, we are able to estimate the size of the secondary lens, 1 - 3 au. Although this an appropriate size for a structure that could cause an extreme scattering event, we do not have conclusive evidence for or against that possibility., Comment: 20 pages, 19 figures

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

Text-to-video generation enhances content creation but is highly computationally intensive: The computational cost of Diffusion Transformers (DiTs) scales quadratically in the number of pixels. This makes minute-length video generation extremely expensive, limiting most existing models to generating videos of only 10-20 seconds length. We propose a Linear-complexity text-to-video Generation (LinGen) framework whose cost scales linearly in the number of pixels. For the first time, LinGen enables high-resolution minute-length video generation on a single GPU without compromising quality. It replaces the computationally-dominant and quadratic-complexity block, self-attention, with a linear-complexity block called MATE, which consists of an MA-branch and a TE-branch. The MA-branch targets short-to-long-range correlations, combining a bidirectional Mamba2 block with our token rearrangement method, Rotary Major Scan, and our review tokens developed for long video generation. The TE-branch is a novel TEmporal Swin Attention block that focuses on temporal correlations between adjacent tokens and medium-range tokens. The MATE block addresses the adjacency preservation issue of Mamba and improves the consistency of generated videos significantly. Experimental results show that LinGen outperforms DiT (with a 75.6% win rate) in video quality with up to 15$\times$ (11.5$\times$) FLOPs (latency) reduction. Furthermore, both automatic metrics and human evaluation demonstrate our LinGen-4B yields comparable video quality to state-of-the-art models (with a 50.5%, 52.1%, 49.1% win rate with respect to Gen-3, LumaLabs, and Kling, respectively). This paves the way to hour-length movie generation and real-time interactive video generation. We provide 68s video generation results and more examples in our project website: https://lineargen.github.io/., Comment: 20 pages, 20 figures

In this paper, we introduce Quantum-Train-Based Distributed Multi-Agent Reinforcement Learning (Dist-QTRL), a novel approach to addressing the scalability challenges of traditional Reinforcement Learning (RL) by integrating quantum computing principles. Quantum-Train Reinforcement Learning (QTRL) leverages parameterized quantum circuits to efficiently generate neural network parameters, achieving a \(poly(\log(N))\) reduction in the dimensionality of trainable parameters while harnessing quantum entanglement for superior data representation. The framework is designed for distributed multi-agent environments, where multiple agents, modeled as Quantum Processing Units (QPUs), operate in parallel, enabling faster convergence and enhanced scalability. Additionally, the Dist-QTRL framework can be extended to high-performance computing (HPC) environments by utilizing distributed quantum training for parameter reduction in classical neural networks, followed by inference using classical CPUs or GPUs. This hybrid quantum-HPC approach allows for further optimization in real-world applications. In this paper, we provide a mathematical formulation of the Dist-QTRL framework and explore its convergence properties, supported by empirical results demonstrating performance improvements over centric QTRL models. The results highlight the potential of quantum-enhanced RL in tackling complex, high-dimensional tasks, particularly in distributed computing settings, where our framework achieves significant speedups through parallelization without compromising model accuracy. This work paves the way for scalable, quantum-enhanced RL systems in practical applications, leveraging both quantum and classical computational resources.