Search

The deployment of advanced deep learning models for medical image segmentation is often constrained by the requirement for extensively annotated datasets. Weakly-supervised learning, which allows less precise labels, has become a promising solution to this challenge. Building on this approach, we propose QMaxViT-Unet+, a novel framework for scribble-supervised medical image segmentation. This framework is built on the U-Net architecture, with the encoder and decoder replaced by Multi-Axis Vision Transformer (MaxViT) blocks. These blocks enhance the model's ability to learn local and global features efficiently. Additionally, our approach integrates a query-based Transformer decoder to refine features and an edge enhancement module to compensate for the limited boundary information in the scribble label. We evaluate the proposed QMaxViT-Unet+ on four public datasets focused on cardiac structures, colorectal polyps, and breast cancer: ACDC, MS-CMRSeg, SUN-SEG, and BUSI. Evaluation metrics include the Dice similarity coefficient (DSC) and the 95th percentile of Hausdorff distance (HD95). Experimental results show that QMaxViT-Unet+ achieves 89.1\% DSC and 1.316mm HD95 on ACDC, 88.4\% DSC and 2.226mm HD95 on MS-CMRSeg, 71.4\% DSC and 4.996mm HD95 on SUN-SEG, and 69.4\% DSC and 50.122mm HD95 on BUSI. These results demonstrate that our method outperforms existing approaches in terms of accuracy, robustness, and efficiency while remaining competitive with fully-supervised learning approaches. This makes it ideal for medical image analysis, where high-quality annotations are often scarce and require significant effort and expense. The code is available at: https://github.com/anpc849/QMaxViT-Unet

Identifying cause-and-effect relationships is critical to understanding real-world dynamics and ultimately causal reasoning. Existing methods for identifying event causality in NLP, including those based on Large Language Models (LLMs), exhibit difficulties in out-of-distribution settings due to the limited scale and heavy reliance on lexical cues within available benchmarks. Modern benchmarks, inspired by probabilistic causal inference, have attempted to construct causal graphs of events as a robust representation of causal knowledge, where \texttt{CRAB} \citep{romanou2023crab} is one such recent benchmark along this line. In this paper, we introduce \texttt{ACCESS}, a benchmark designed for discovery and reasoning over abstract causal events. Unlike existing resources, \texttt{ACCESS} focuses on causality of everyday life events on the abstraction level. We propose a pipeline for identifying abstractions for event generalizations from \texttt{GLUCOSE} \citep{mostafazadeh-etal-2020-glucose}, a large-scale dataset of implicit commonsense causal knowledge, from which we subsequently extract $1,4$K causal pairs. Our experiments highlight the ongoing challenges of using statistical methods and/or LLMs for automatic abstraction identification and causal discovery in NLP. Nonetheless, we demonstrate that the abstract causal knowledge provided in \texttt{ACCESS} can be leveraged for enhancing QA reasoning performance in LLMs.

As Large Language Models (LLMs) evolve from text-completion tools into fully fledged agents operating in dynamic environments, they must address the challenge of continually learning and retaining long-term knowledge. Many biological systems solve these challenges with episodic memory, which supports single-shot learning of instance-specific contexts. Inspired by this, we present an episodic memory framework for LLM agents, centered around five key properties of episodic memory that underlie adaptive and context-sensitive behavior. With various research efforts already partially covering these properties, this position paper argues that now is the right time for an explicit, integrated focus on episodic memory to catalyze the development of long-term agents. To this end, we outline a roadmap that unites several research directions under the goal to support all five properties of episodic memory for more efficient long-term LLM agents.

Implementing carbon capture technology on-board ships holds promise as a solution to facilitate the reduction of carbon intensity in international shipping, as mandated by the International Maritime Organization. In this work, we address the energy-efficient operation of shipboard carbon capture processes by proposing a hybrid modeling-based economic predictive control scheme. Specifically, we consider a comprehensive shipboard carbon capture process that encompasses the ship engine system and the shipboard post-combustion carbon capture plant. To accurately and robustly characterize the dynamic behaviors of this shipboard plant, we develop a hybrid dynamic process model that integrates available imperfect physical knowledge with neural networks trained using process operation data. An economic model predictive control approach is proposed based on the hybrid model to ensure carbon capture efficiency while minimizing energy consumption required for the carbon capture process operation. \textcolor{blue}{The cross-entropy method is employed to efficiently solve the complex non-convex optimization problem associated with the proposed hybrid model-based economic model predictive control method.} Extensive simulations, analyses, and comparisons are conducted to verify the effectiveness and illustrate the superiority of the proposed framework., Comment: 56 pages, 21 figures, 12 tables

We investigate algorithmic decision problems where agents can respond strategically to the decision maker's (DM) models. The demand for clear and actionable explanations from DMs to (potentially strategic) agents continues to rise. While prior work often treats explanations as full model disclosures, explanations in practice might convey only partial information, which can lead to misinterpretations and harmful responses. When full disclosure of the predictive model is neither feasible nor desirable, a key open question is how DMs can use explanations to maximise their utility without compromising agent welfare. In this work, we explore well-known local and global explanation methods, and establish a necessary condition to prevent explanations from misleading agents into self-harming actions. Moreover, with conditional homogeneity, we establish that action recommendation (AR)-based explanations are sufficient for non-harmful responses, akin to the revelation principle in information design. To operationalise AR-based explanations, we propose a simple algorithm to jointly optimise the predictive model and AR policy to balance DM outcomes with agent welfare. Our empirical results demonstrate the benefits of this approach as a more refined strategy for safe and effective partial model disclosure in algorithmic decision-making.

Contrastive learning is a well-established paradigm in representation learning. The standard framework of contrastive learning minimizes the distance between "similar" instances and maximizes the distance between dissimilar ones in the projection space, disregarding the various aspects of similarity that can exist between two samples. Current methods rely on a single projection head, which fails to capture the full complexity of different aspects of a sample, leading to suboptimal performance, especially in scenarios with limited training data. In this paper, we present a novel supervised contrastive learning method in a unified framework called multilevel contrastive learning (MLCL), that can be applied to both multi-label and hierarchical classification tasks. The key strength of the proposed method is the ability to capture similarities between samples across different labels and/or hierarchies using multiple projection heads. Extensive experiments on text and image datasets demonstrate that the proposed approach outperforms state-of-the-art contrastive learning methods

The analytic expressions for one-loop contributions to the rare decay process $A_0 \rightarrow h_0h_0\gamma$ within the CP-conserving of Two Higgs Doublet Models are first reported in this paper. Analytic results are presented in term of scalar one-loop Passarino-Veltman functions following the standard output of the packages~{\tt LoopTools} and {\tt Collier}. In this context, physical results for the computed process are easily generated by using one of these packages. The numerical checks are proposed to verify for the analytic results in this paper. The checks rely on the renormalization conditions that the decay amplitude must be the ultraviolet finiteness and infrared finiteness. The amplitude consisting of an external photon always obeys the Ward identity. This will be confirmed numerically in this article. In phenomenological results, the decay rates of $A_0 \rightarrow h_0h_0\gamma$ are evaluated at several points in the allowed regions of the parameter space. Furthermore, the differential decay widths with respect to the invariant mass of Higgs-pair in final states are studied., Comment: 30 pages, 4 Tables, 6 Figures

Ensuring the correctness of code generated by Large Language Models (LLMs) presents a significant challenge in AI-driven software development. Existing approaches predominantly rely on black-box (closed-box) approaches that evaluate correctness post-generation, failing to utilize the rich insights embedded in the LLMs' internal states during code generation. In this paper, we introduce OPENIA, a novel white-box (open-box) framework that leverages these internal representations to assess the correctness of LLM-generated code. OPENIA systematically analyzes the intermediate states of representative open-source LLMs specialized for code, including DeepSeek-Coder, CodeLlama, and MagicCoder, across diverse code generation benchmarks. Our empirical analysis reveals that these internal representations encode latent information, which strongly correlates with the correctness of the generated code. Building on these insights, OPENIA uses a white-box/open-box approach to make informed predictions about code correctness, offering significant advantages in adaptability and robustness over traditional classification-based methods and zero-shot approaches. Experimental results demonstrate that OPENIA consistently outperforms baseline models, achieving higher accuracy, precision, recall, and F1-Scores with up to a 2X improvement in standalone code generation and a 46% enhancement in repository-specific scenarios. By unlocking the potential of in-process signals, OPENIA paves the way for more proactive and efficient quality assurance mechanisms in LLM-assisted code generation.

Let $\mathfrak A$ denote the classical singly-graded Steenrod algebra over the binary field $\mathbb Z/2.$ We write $P_k:=\mathbb Z/2[t_1, t_2, \ldots, t_k]$ as the polynomial algebra on $k$ generators, each having a degree of one. Let $GL_k$ be the general linear group of rank $k$ over $\mathbb Z/2.$ Then, $P_k$ is an $ \mathfrak A[GL_k]$-module. The structure of the cohomology groups, ${\rm Ext}_{ \mathfrak A}^{k, k+\bullet}(\mathbb Z/2, \mathbb Z/2)$, of the Steenrod algebra has, thus far, resisted clear understanding and full description for all homological degrees $k$. In the study of these groups, the algebraic transfer$-$constructed by W. Singer in [Math. Z. 202, 493--523 (1989)]$-$plays an important role. The Singer transfer is represented by the following homomorphism: $$Tr_k: {\rm Hom}([(\mathbb Z/2\otimes_{ \mathfrak A} P_k)_{\bullet}]^{GL_k}, \mathbb Z/2)\longrightarrow {\rm Ext}_{ \mathfrak A}^{k, k+\bullet}(\mathbb Z/2, \mathbb Z/2).$$ Among Singer's contributions is an interesting open conjecture asserting the monomorphism of $Tr_k$ for all $k.$ For this reason, our main aim in this article is to ascertain the validity of the Singer conjecture for ranks 4 and 6 in certain families of internal degrees. We place particular emphasis on the rank 4 case. More precisely, we present a detailed proof for certain generic degree cases when verifying the conjecture of rank four, which were succinctly noted in our previous work [Proc. Roy. Soc. Edinburgh Sect. A 153, 1529--1542 (2023)]., Comment: This article consists of 36 pages, written in French

The covariance matrix is a foundation in numerous statistical and machine-learning applications such as Principle Component Analysis, Correlation Heatmap, etc. However, missing values within datasets present a formidable obstacle to accurately estimating this matrix. While imputation methods offer one avenue for addressing this challenge, they often entail a trade-off between computational efficiency and estimation accuracy. Consequently, attention has shifted towards direct parameter estimation, given its precision and reduced computational burden. In this paper, we propose Direct Parameter Estimation for Randomly Missing Data with Categorical Features (DPERC), an efficient approach for direct parameter estimation tailored to mixed data that contains missing values within continuous features. Our method is motivated by leveraging information from categorical features, which can significantly enhance covariance matrix estimation for continuous features. Our approach effectively harnesses the information embedded within mixed data structures. Through comprehensive evaluations of diverse datasets, we demonstrate the competitive performance of DPERC compared to various contemporary techniques. In addition, we also show by experiments that DPERC is a valuable tool for visualizing the correlation heatmap.

In high-dimensional regression, feature selection methods, such as sequential feature selection (SeqFS), are commonly used to identify relevant features. When data is limited, domain adaptation (DA) becomes crucial for transferring knowledge from a related source domain to a target domain, improving generalization performance. Although SeqFS after DA is an important task in machine learning, none of the existing methods can guarantee the reliability of its results. In this paper, we propose a novel method for testing the features selected by SeqFS-DA. The main advantage of the proposed method is its capability to control the false positive rate (FPR) below a significance level $\alpha$ (e.g., 0.05). Additionally, a strategic approach is introduced to enhance the statistical power of the test. Furthermore, we provide extensions of the proposed method to SeqFS with model selection criteria including AIC, BIC, and adjusted R-squared. Extensive experiments are conducted on both synthetic and real-world datasets to validate the theoretical results and demonstrate the proposed method's superior performance.

Content providers increasingly utilise Content Delivery Networks (CDNs) to enhance users' content download experience. However, this deployment scenario raises significant security concerns regarding content confidentiality and user privacy due to the involvement of third-party providers. Prior proposals using private information retrieval (PIR) and oblivious RAM (ORAM) have proven impractical due to high computation and communication costs, as well as integration challenges within distributed CDN architectures. In response, we present \textsf{OblivCDN}, a practical privacy-preserving system meticulously designed for seamless integration with the existing real-world Internet-CDN infrastructure. Our design strategically adapts Range ORAM primitives to optimise memory and disk seeks when accessing contiguous blocks of CDN content, both at the origin and edge servers, while preserving both content confidentiality and user access pattern hiding features. Also, we carefully customise several oblivious building blocks that integrate the distributed trust model into the ORAM client, thereby eliminating the computational bottleneck in the origin server and reducing communication costs between the origin server and edge servers. Moreover, the newly-designed ORAM client also eliminates the need for trusted hardware on edge servers, and thus significantly ameliorates the compatibility towards networks with massive legacy devices.In real-world streaming evaluations, OblivCDN} demonstrates remarkable performance, downloading a $256$ MB video in just $5.6$ seconds. This achievement represents a speedup of $90\times$ compared to a strawman approach (direct ORAM adoption) and a $366\times$ improvement over the prior art, OblivP2P., Comment: The 20th ACM ASIA Conference on Computer and Communications Security (ACM ASIACCS 2025)

In this paper, we establish characterizations of variational $s$-convexity and tilt stability for prox-regular functions in the absence of subdifferential continuity via quadratic bundles, a kind of primal-dual generalized second-order derivatives recently introduced by Rockafellar. Deriving such characterizations in the effective pointbased form requires a certain revision of quadratic bundles investigated below. Our device is based on the notion of generalized twice differentiability and its novel characterization via classical twice differentiability of the associated Moreau envelopes combined with various limiting procedures for functions and sets.

While ViTs have achieved across machine learning tasks, deploying them in real-world scenarios faces a critical challenge: generalizing under OoD shifts. A crucial research gap exists in understanding how to design ViT architectures, both manually and automatically, for better OoD generalization. To this end, we introduce OoD-ViT-NAS, the first systematic benchmark for ViTs NAS focused on OoD generalization. This benchmark includes 3000 ViT architectures of varying computational budgets evaluated on 8 common OoD datasets. Using this benchmark, we analyze factors contributing to OoD generalization. Our findings reveal key insights. First, ViT architecture designs significantly affect OoD generalization. Second, ID accuracy is often a poor indicator of OoD accuracy, highlighting the risk of optimizing ViT architectures solely for ID performance. Third, we perform the first study of NAS for ViTs OoD robustness, analyzing 9 Training-free NAS methods. We find that existing Training-free NAS methods are largely ineffective in predicting OoD accuracy despite excelling at ID accuracy. Simple proxies like Param or Flop surprisingly outperform complex Training-free NAS methods in predicting OoD accuracy. Finally, we study how ViT architectural attributes impact OoD generalization and discover that increasing embedding dimensions generally enhances performance. Our benchmark shows that ViT architectures exhibit a wide range of OoD accuracy, with up to 11.85% improvement for some OoD shifts. This underscores the importance of studying ViT architecture design for OoD. We believe OoD-ViT-NAS can catalyze further research into how ViT designs influence OoD generalization., Comment: Accepted in NeurIPS 2024

The minimal Composite Higgs model (MCHM) provides an effective trigger for the Baryogenesis scenario through the confinement scale of the strong sector ($f$) or Dilaton ($\chi$). $f$ is a parameter with mass dimension, which stores the resonances of particles at high energies, has a suitable value of about 800 GeV. But when $300$ GeV $\le f \le 400$ GeV, the effective Higgs potential has a first order electroweak phase transition. Therefore, although $f$ cannot be a perfect trigger, it does suggest an effective approach that accommodates the resonances of particles. Thus the investigation of the electroweak phase transition according to $f$ has confirmed that the inclusion of Dilaton in the effective potential is reasonable. Accordingly, we derive a Dilaton potential with appropriate parameter domains and $f=800$ GeV, the mass of Dilaton ranges from 300 GeV to 700 GeV, which will give an electroweak phase transition strength greater than 1 and less than 3, enough for a first order phase transition. This is a directly and clearly prove of the triggers for the first order EWPT in MCHM., Comment: 27 pages, 4 figures

In this paper, we investigate the concepts of generalized twice differentiability and quadratic bundles of nonsmooth functions that have been very recently proposed by Rockafellar in the framework of second-order variational analysis. These constructions, in contrast to second-order subdifferentials, are defined in primal spaces. We develop new techniques to study generalized twice differentiability for a broad class of prox-regular functions, establish their novel characterizations. Subsequently, quadratic bundles of prox-regular functions are shown to be nonempty, which provides the ground of potential applications in variational analysis and optimization.

Nonlinear differential equations and systems play a crucial role in modeling systems where time-dependent factors exhibit nonlinear characteristics. Due to their nonlinear nature, solving such systems often presents significant difficulties and challenges. In this study, we propose a method utilizing Physics-Informed Neural Networks (PINNs) to solve the nonlinear energy supply-demand (ESD) system. We design a neural network with four outputs, where each output approximates a function that corresponds to one of the unknown functions in the nonlinear system of differential equations describing the four-dimensional ESD problem. The neural network model is then trained and the parameters are identified, optimized to achieve a more accurate solution. The solutions obtained from the neural network for this problem are equivalent when we compare and evaluate them against the Runge-Kutta numerical method of order 4/5 (RK45). However, the method utilizing neural networks is considered a modern and promising approach, as it effectively exploits the superior computational power of advanced computer systems, especially in solving complex problems. Another advantage is that the neural network model, after being trained, can solve the nonlinear system of differential equations across a continuous domain. In other words, neural networks are not only trained to approximate the solution functions for the nonlinear ESD system but can also represent the complex dynamic relationships between the system's components. However, this approach requires significant time and computational power due to the need for model training., Comment: Submitted to Computation J

Self-supervised visual foundation models produce powerful embeddings that achieve remarkable performance on a wide range of downstream tasks. However, unlike vision-language models such as CLIP, self-supervised visual features are not readily aligned with language, hindering their adoption in open-vocabulary tasks. Our method, named dino.txt, unlocks this new ability for DINOv2, a widely used self-supervised visual encoder. We build upon the LiT training strategy, which trains a text encoder to align with a frozen vision model but leads to unsatisfactory results on dense tasks. We propose several key ingredients to improve performance on both global and dense tasks, such as concatenating the [CLS] token with the patch average to train the alignment and curating data using both text and image modalities. With these, we successfully train a CLIP-like model with only a fraction of the computational cost compared to CLIP while achieving state-of-the-art results in zero-shot classification and open-vocabulary semantic segmentation.

Recovering a low-complexity signal from its noisy observations by regularization methods is a cornerstone of inverse problems and compressed sensing. Stable recovery ensures that the original signal can be approximated linearly by optimal solutions of the corresponding Morozov or Tikhonov regularized optimization problems. In this paper, we propose new characterizations for stable recovery in finite-dimensional spaces, uncovering the role of nonsmooth second-order information. These insights enable a deeper understanding of stable recovery and their practical implications. As a consequence, we apply our theory to derive new sufficient conditions for stable recovery of the analysis group sparsity problems, including the group sparsity and isotropic total variation problems. Numerical experiments on these two problems give favorable results about using our conditions to test stable recovery., Comment: 29 pages, 4 figures

Background: This protocol introduces an ultrasound (US) configuration for whole-body 3D dynamic contrast-enhanced ultrasound (DCE-US) imaging in preclinical applications. The set-up relies on a clinical abdominal matrix US probe to enable mice imaging beyond current preclinical systems that are generally unable to capture whole-body volumetric and dynamic imaging. We demonstrate via this protocol the feasibility of volumetric contrast-enhanced and molecular 3D imaging in the entire lower body of mouse, as well as the capability of imaging multiple lesions in the same animal simultaneously with a single contrast bolus injection. Methods: We modified a silicone cup with a 101.6 mm inner diameter and 6.4 mm wall thickness, to a height of 76.2 mm and cut out a rectangular side window (12.7 mm x 15.9 mm) at the lower part of the cup. Mice were positioned with their head and the two front legs outside of the cup while the rest of the body remained inside the cup and submerged under water; the edges around the mouse's body in contact with the cup's wall were sealed with Vaseline to prevent water leakage, and the cup was filled with warm water to maintain the mouse's body heat. Imaging was conducted using a portable clinical ultrasound system (EPIQ 7 Philips) with an abdominal matrix-array transducer (PM mode, X6-1) inserted into the top of the cup, and positioned to visualize the entire of the mouse's abdomen. The mice were imaged before and after the injection of P-selectin-targeted microbubbles. Results: Here, in nude mice with two adjacent tumors, we demonstrate that a commercially available clinical matrix transducer can be utilized to achieve whole-body 3D DCE-US and molecular US imaging, as well as capture independent qualitative and quantitative information from several lesions or organs at different locations in the animal, through a single bolus injection., Comment: 16 pages, 4 figures

Low-light and blurring issues are prevalent when capturing photos at night, often due to the use of long exposure to address dim environments. Addressing these joint problems can be challenging and error-prone if an end-to-end model is trained without incorporating an appropriate physical model. In this paper, we introduce JUDE, a Deep Joint Unrolling for Deblurring and Low-Light Image Enhancement, inspired by the image physical model. Based on Retinex theory and the blurring model, the low-light blurry input is iteratively deblurred and decomposed, producing sharp low-light reflectance and illuminance through an unrolling mechanism. Additionally, we incorporate various modules to estimate the initial blur kernel, enhance brightness, and eliminate noise in the final image. Comprehensive experiments on LOL-Blur and Real-LOL-Blur demonstrate that our method outperforms existing techniques both quantitatively and qualitatively., Comment: 10 pages

Reconstructing high-fidelity fluid flow fields from sparse sensor measurements is vital for many science and engineering applications but remains challenging because of dimensional disparities between state and observational spaces. Due to such dimensional differences, the measurement operator becomes ill-conditioned and non-invertible, making the reconstruction of flow fields from sensor measurements extremely difficult. Although sparse optimization and machine learning address the above problems to some extent, questions about their generalization and efficiency remain, particularly regarding the discretization dependence of these models. In this context, deep operator learning offers a better solution as this approach models mappings between infinite-dimensional functional spaces, enabling superior generalization and discretization-independent reconstruction. We introduce FLRONet, a deep operator learning framework that is trained to reconstruct fluid flow fields from sparse sensor measurements. FLRONet employs a branch-trunk network architecture to represent the inverse measurement operator that maps sensor observations to the original flow field, a continuous function of both space and time. Validation performed on the CFDBench dataset has demonstrated that FLRONet consistently achieves high levels of reconstruction accuracy and robustness, even in scenarios where sensor measurements are inaccurate or missing. Furthermore, the operator learning approach endows FLRONet with the capability to perform zero-shot super-resolution in both spatial and temporal domains, offering a solution for rapid reconstruction of high-fidelity flow fields.

As Large Language Models (LLMs) scale to longer context windows, the computational cost of attention mechanisms, which traditionally grows quadratically with input length, presents a critical challenge for real-time and memory-constrained deployments. Existing sparse attention techniques have sought to reduce this complexity, but they often incur significant overhead or compromise accuracy, making them less practical for large contexts on mid-range hardware. In this paper, we introduce SparseAccelerate, a dynamic sparse attention method that adapts its sparsity patterns based on input characteristics, effectively flattening the attention complexity curve. Our approach is effective for input lengths starting at 16K tokens and scales efficiently up to 128K tokens on dual NVIDIA A5000 GPUs (24GB each). Experimental results show that SparseAccelerate achieves up to a 1.04x reduction in Time-To-First-Token (TTFT) latency at 32K tokens, while also providing substantial memory savings. These improvements yield practical gains for memory-intensive applications and long-context tasks that were previously infeasible with standard attention. Beyond latency reductions, SparseAccelerate fundamentally shifts the scaling trend, demonstrating the smallest TTFT growth gradient relative to context length among competing methods. Ongoing evaluations on diverse benchmarks confirm its scalability, positioning SparseAccelerate as a critical advancement toward efficient, real-time, and large-context LLM inference on accessible hardware.

Text Retrieval (TR) involves finding and retrieving text-based content relevant to a user's query from a large repository, with applications in real-world scenarios such as legal document retrieval. While most existing studies focus on English, limited work addresses Japanese contexts. In this paper, we introduce a new dataset specifically designed for Japanese legal contexts and propose a novel two-phase pipeline tailored to this domain. In the first phase, the model learns a broad understanding of global contexts, enhancing its generalization and adaptability to diverse queries. In the second phase, the model is fine-tuned to address complex queries specific to legal scenarios. Extensive experiments are conducted to demonstrate the superior performance of our method, which outperforms existing baselines. Furthermore, our pipeline proves effective in English contexts, surpassing comparable baselines on the MS MARCO dataset. We have made our code publicly available on GitHub, and the model checkpoints are accessible via HuggingFace.

Autonomous vehicles (AVs) have demonstrated significant potential in revolutionizing transportation, yet ensuring their safety and reliability remains a critical challenge, especially when exposed to dynamic and unpredictable environments. Real-world testing of an Autonomous Driving System (ADS) is both expensive and risky, making simulation-based testing a preferred approach. In this paper, we propose AVASTRA, a Reinforcement Learning (RL)-based approach to generate realistic critical scenarios for testing ADSs in simulation environments. To capture the complexity of driving scenarios, AVASTRA comprehensively represents the environment by both the internal states of an ADS under-test (e.g., the status of the ADS's core components, speed, or acceleration) and the external states of the surrounding factors in the simulation environment (e.g., weather, traffic flow, or road condition). AVASTRA trains the RL agent to effectively configure the simulation environment that places the AV in dangerous situations and potentially leads it to collisions. We introduce a diverse set of actions that allows the RL agent to systematically configure both environmental conditions and traffic participants. Additionally, based on established safety requirements, we enforce heuristic constraints to ensure the realism and relevance of the generated test scenarios. AVASTRA is evaluated on two popular simulation maps with four different road configurations. Our results show AVASTRA's ability to outperform the state-of-the-art approach by generating 30% to 115% more collision scenarios. Compared to the baseline based on Random Search, AVASTRA achieves up to 275% better performance. These results highlight the effectiveness of AVASTRA in enhancing the safety testing of AVs through realistic comprehensive critical scenario generation.

This paper investigates Singer's conjecture by examining the cohit module $\mathbb F_2\otimes_{\mathcal A}P^{\otimes h}$ for specific degrees and values of $h$. Utilizing hit problem techniques, we extend previous work by Mothebe et al. and establish key dimensional results. Notably, for $h\geq 6$, we prove that the cohit module's dimension in certain degrees matches the order of a specific factor group. Our contributions include demonstrating that certain non-zero elements do not belong to the image of the Singer algebraic transfer., Comment: 37 pages

Test-time adaptation (TTA) has emerged as a promising solution to tackle the continual domain shift in machine learning by allowing model parameters to change at test time, via self-supervised learning on unlabeled testing data. At the same time, it unfortunately opens the door to unforeseen vulnerabilities for degradation over time. Through a simple theoretical continual TTA model, we successfully identify a risk in the sampling process of testing data that could easily degrade the performance of a continual TTA model. We name this risk as Reusing of Incorrect Prediction (RIP) that TTA attackers can employ or as a result of the unintended query from general TTA users. The risk posed by RIP is also highly realistic, as it does not require prior knowledge of model parameters or modification of testing samples. This simple requirement makes RIP as the first black-box TTA attack algorithm that stands out from existing white-box attempts. We extensively benchmark the performance of the most recent continual TTA approaches when facing the RIP attack, providing insights on its success, and laying out potential roadmaps that could enhance the resilience of future continual TTA systems.

This article presents the Labeled Random Finite Set (LRFS) framework for multi-object systems-systems in which the number of objects and their states are unknown and vary randomly with time. In particular, we focus on state and trajectory estimation via a multi-object State Space Model (SSM) that admits principled tractable multi-object tracking filters/smoothers. Unlike the single-object counterpart, a time sequence of states does not necessarily represent the trajectory of a multi-object system. The LRFS formulation enables a time sequence of multi-object states to represent the multi-object trajectory that accommodates trajectory crossings and fragmentations. We present the basics of LRFS, covering a suite of commonly used models and mathematical apparatus (including the latest results not published elsewhere). Building on this, we outline the fundamentals of multi-object state space modeling and estimation using LRFS, which formally address object identities/trajectories, ancestries for spawning objects, and characterization of the uncertainty on the ensemble of objects (and their trajectories). Numerical solutions to multi-object SSM problems are inherently far more challenging than those in standard SSM. To bridge the gap between theory and practice, we discuss state-of-the-art implementations that address key computational bottlenecks in the number of objects, measurements, sensors, and scans.

Multimodal Large Language Models (MLLMs) have made notable advances in visual understanding, yet their abilities to recognize objects modified by specific attributes remain an open question. To address this, we explore MLLMs' reasoning capabilities in object recognition, ranging from commonsense to beyond-commonsense scenarios. We introduce a novel benchmark, NEMO, which comprises 900 images of origiNal fruits and their corresponding attributE-MOdified ones; along with a set of 2,700 questions including open-, multiple-choice-, unsolvable types. We assess 26 recent open-sourced and commercial models using our benchmark. The findings highlight pronounced performance gaps in recognizing objects in NEMO and reveal distinct answer preferences across different models. Although stronger vision encoders improve performance, MLLMs still lag behind standalone vision encoders. Interestingly, scaling up the model size does not consistently yield better outcomes, as deeper analysis reveals that larger LLMs can weaken vision encoders during fine-tuning. These insights shed light on critical limitations in current MLLMs and suggest potential pathways toward developing more versatile and resilient multimodal models.

Multi-user multiple-input multiple-output (MU-MIMO) systems allow multiple users to share the same wireless spectrum. Each user transmits one symbol drawn from an M-ary quadrature amplitude modulation (M-QAM) constellation set, and the resulting multi-user interference (MUI) is cancelled at the receiver. End-to-End (E2E) learning has recently been proposed to jointly design the constellation set for a modulator and symbol detector for a single user under an additive white Gaussian noise (AWGN) channel by using deep neural networks (DNN) with a symbol-error-rate (SER) performance, exceeding that of Maximum likelihood (ML) M-QAM detectors. In this paper, we extend the E2E concept to the MU-MIMO systems, where a DNN-based modulator that generates learned M constellation points and graph expectation propagation network (GEPNet) detector that cancels MUI are jointly optimised with respect to SER performance loss. Simulation results demonstrate that the proposed E2E with learned constellation outperforms GEPNet with 16-QAM by around 5 dB in terms of SER in a high MUI environment and even surpasses ML with 16-QAM in a low MUI condition, both with no additional computational complexity., Comment: submitted to a conference

Bandgaps in layered materials are critical for enabling functionalities such as tunable photodetection, efficient energy conversion, and nonlinear optical responses, which are essential for next-generation photonic and quantum devices. Gap engineering could form heterostructures with complementary materials like transition metal dichalcogenides or perovskites for multi-functional devices. Graphite, conventionally regarded as a gapless material, exhibits a bandgap of ~100 meV in nano-scale patterned highly oriented pyrolytic graphite (HOPG), as revealed by angle-resolved photoemission spectroscopy (ARPES) and Raman measurements. Our state-of-the-art calculations, incorporating photoemission matrix element effects, predict this bandgap with remarkable accuracy and attribute it to mechanical distortions introduced during patterning. This work bridges theory and experiment, providing the direct evidence of a tunable bandgap in HOPG. Beyond its fundamental significance, this finding opens new possibilities for designing materials with tailored electronic properties, enabling advancements in terahertz devices and optoelectronics., Comment: 33 pages, 9 figures

Here, we present the outcomes from the second Large Language Model (LLM) Hackathon for Applications in Materials Science and Chemistry, which engaged participants across global hybrid locations, resulting in 34 team submissions. The submissions spanned seven key application areas and demonstrated the diverse utility of LLMs for applications in (1) molecular and material property prediction; (2) molecular and material design; (3) automation and novel interfaces; (4) scientific communication and education; (5) research data management and automation; (6) hypothesis generation and evaluation; and (7) knowledge extraction and reasoning from scientific literature. Each team submission is presented in a summary table with links to the code and as brief papers in the appendix. Beyond team results, we discuss the hackathon event and its hybrid format, which included physical hubs in Toronto, Montreal, San Francisco, Berlin, Lausanne, and Tokyo, alongside a global online hub to enable local and virtual collaboration. Overall, the event highlighted significant improvements in LLM capabilities since the previous year's hackathon, suggesting continued expansion of LLMs for applications in materials science and chemistry research. These outcomes demonstrate the dual utility of LLMs as both multipurpose models for diverse machine learning tasks and platforms for rapid prototyping custom applications in scientific research., Comment: Updating author information, the submission remains largely unchanged. 98 pages total

Large Language Models (LLMs) have demonstrated the potential to address some issues within the semiconductor industry. However, they are often general-purpose models that lack the specialized knowledge needed to tackle the unique challenges of this sector, such as the intricate physics and chemistry of semiconductor devices and processes. SemiKong, the first industry-specific LLM for the semiconductor domain, provides a foundation that can be used to develop tailored proprietary models. With SemiKong 1.0, we aim to develop a foundational model capable of understanding etching problems at an expert level. Our key contributions include (a) curating a comprehensive corpus of semiconductor-related texts, (b) creating a foundational model with in-depth semiconductor knowledge, and (c) introducing a framework for integrating expert knowledge, thereby advancing the evaluation process of domain-specific AI models. Through fine-tuning a pre-trained LLM using our curated dataset, we have shown that SemiKong outperforms larger, general-purpose LLMs in various semiconductor manufacturing and design tasks. Our extensive experiments underscore the importance of developing domain-specific LLMs as a foundation for company- or tool-specific proprietary models, paving the way for further research and applications in the semiconductor domain. Code and dataset will be available at https://github.com/aitomatic/semikong, Comment: On-going work

Naturalistic driving action localization task aims to recognize and comprehend human behaviors and actions from video data captured during real-world driving scenarios. Previous studies have shown great action localization performance by applying a recognition model followed by probability-based post-processing. Nevertheless, the probabilities provided by the recognition model frequently contain confused information causing challenge for post-processing. In this work, we adopt an action recognition model based on self-supervise learning to detect distracted activities and give potential action probabilities. Subsequently, a constraint ensemble strategy takes advantages of multi-camera views to provide robust predictions. Finally, we introduce a conditional post-processing operation to locate distracted behaviours and action temporal boundaries precisely. Experimenting on test set A2, our method obtains the sixth position on the public leaderboard of track 3 of the 2024 AI City Challenge., Comment: Computer Vision and Pattern Recognition Workshop 2024

In this paper, we propose ZeFaV - a zero-shot based fact-checking verification framework to enhance the performance on fact verification task of large language models by leveraging the in-context learning ability of large language models to extract the relations among the entities within a claim, re-organized the information from the evidence in a relationally logical form, and combine the above information with the original evidence to generate the context from which our fact-checking model provide verdicts for the input claims. We conducted empirical experiments to evaluate our approach on two multi-hop fact-checking datasets including HoVer and FEVEROUS, and achieved potential results results comparable to other state-of-the-art fact verification task methods., Comment: This pre-print has been published in PRICAI 2024: Trends in Artificial Intelligence. The published version is available at https://doi.org/10.1007/978-981-96-0119-6_28

Photoelectron diffraction (PED) is a powerful and essential experimental technique for resolving the structure of surfaces with sub-angstrom resolution. In the high energy regime, researchers in angle-resolved photoemission spectroscopy (ARPES) observe modulating patterns attributed to X-ray-PED (XPD) effects. This is accompanied by other challenges such as low cross-sections, significant photon momentum transfer, and non-negligible phonon scattering. Overall, XPD is not only an advantageous approach but also exhibits unexpected effects. To disentangle these diffraction influences, we present a PED implementation for the SPRKKR package that utilizes multiple scattering theory and a one-step model in the photoemission process. Unlike real-space implementations of the multiple scattering XPD formalism, we propose a k-space implementation based on the layer KKR method. The main advantage of this method is its ability to address a very broad kinetic energy range (20-8000 eV) without convergence problems related to angular momentum and cluster size. Furthermore, the so-called alloy analogy model can be used to simulate XPD at finite temperatures as well as XPD effects observed in soft and hard X-ray ARPES. For practical applications, we have calculated the circular dichroism in angular distributions (CDAD) associated with core-level photoemission of 2p from Si(100) and 3p from Ge(100). Photoelectrons are excited by hard X-rays (6000 eV) with right and left circularly polarized radiation (RCP and LCP, respectively).

Retrieval-augmented generation (RAG) is a promising method for addressing some of the memory-related challenges associated with Large Language Models (LLMs). Two separate systems form the RAG pipeline, the retriever and the reader, and the impact of each on downstream task performance is not well-understood. Here, we work towards the goal of understanding how retrievers can be optimized for RAG pipelines for common tasks such as Question Answering (QA). We conduct experiments focused on the relationship between retrieval and RAG performance on QA and attributed QA and unveil a number of insights useful to practitioners developing high-performance RAG pipelines. For example, lowering search accuracy has minor implications for RAG performance while potentially increasing retrieval speed and memory efficiency., Comment: Accepted to NeurIPS 2024 Workshop ATTRIB

Polariton lasing is a promising phenomenon with potential applications in next-generation lasers that operate without the need for population inversion. Applying a perpendicular magnetic field to a quantum well (QW) significantly alters the properties of exciton-polaritons. In this theoretical study, we investigate how the lasing threshold of QW exciton-polaritons depends on the magnetic field. By modifying the exciton's effective mass and Rabi splitting, the magnetic field induces notable changes in the relaxation kinetics, which directly affect the lasing threshold. For low-energy pumping, an increase in the magnetic field delays the lasing threshold, while for high-energy pumping, the threshold is reached at much lower pump intensities. Furthermore, increasing both the pump energy and the magnetic field enhances relaxation efficiency, leading to a substantially larger number of condensed polaritons. Our result gives insights into the modulation of exciton-polariton condensation through magnetic fields, with potential implications for the design of low-threshold polariton lasers., Comment: 5 pages, 3 figures

We extend the Wirtinger number of links, an invariant originally defined by Blair et al. in terms of extending initial colorings of some strands of a diagram to the entire diagram, to spatial graphs. We prove that the Wirtinger number equals the bridge index of spatial graphs, and we implement an algorithm in Python which gives a more efficient way to estimate upper bounds of bridge indices. Combined with lower bounds from diagram colorings by elements from certain algebraic structures, we obtain exact bridge indices for a large family of spatial graphs., Comment: 27 pages, 21 figures, comments very welcome!

While smart contracts are foundational elements of blockchain applications, their inherent susceptibility to security vulnerabilities poses a significant challenge. Existing training datasets employed for vulnerability detection tools may be limited, potentially compromising their efficacy. This paper presents a method for improving the quantity and quality of smart contract vulnerability datasets and evaluates current detection methods. The approach centers around semantic-preserving code transformation, a technique that modifies the source code structure without altering its semantic meaning. The transformed code snippets are inserted into all potential locations within benign smart contract code, creating new vulnerable contract versions. This method aims to generate a wider variety of vulnerable codes, including those that can bypass detection by current analysis tools. The paper experiments evaluate the method's effectiveness using tools like Slither, Mythril, and CrossFuzz, focusing on metrics like the number of generated vulnerable samples and the false negative rate in detecting these vulnerabilities. The improved results show that many newly created vulnerabilities can bypass tools and the false reporting rate goes up to 100% and increases dataset size minimum by 2.5X.

Long-range imaging in the thermal infrared band is critical for applications such as environmental monitoring, industrial inspections, and surveillance. To achieve high quality imaging, these systems typically require large apertures and many elements with complex shapes to correct aberrations, adding significant weight and cost. Large-area metasurface optics offer a promising solution for weight reduction; however, their substantial chromatic aberrations limit their effectiveness in the long-wave infrared (LWIR) band where broadband imaging is typically desired. In this work, we introduce a hybrid system comprising four refractive lenses and two all-silicon metasurface correctors (meta-correctors) to achieve high-quality, broadband thermal imaging at long range. Compared to a refractive-only assembly, our system demonstrates a three-fold contrast enhancement at the detector's half-Nyquist frequency. Testing outside the laboratory reveals noticeably sharper images, with human features clearly recognizable at distances of 250 meters. The assembly utilizes off-the-shelf refractive elements and avoids the use of germanium, which poses a supply chain risk. Our findings highlight the potential of hybrid meta-corrector systems to enable long-range, lightweight, and cost-effective LWIR imaging solutions.

REST (Representational State Transfer) APIs have become integral for data communication and exchange due to their simplicity, scalability, and compatibility with web standards. However, ensuring REST APIs' reliability through rigorous testing poses significant challenges, given the complexities of operations, parameters, inputs, dependencies, and call sequences. In this paper, we introduce MUCOREST, a novel Reinforcement Learning (RL)-based API testing approach that leverages Q-learning to maximize code coverage and output coverage, thereby improving bug discovery. By focusing on these proximate objectives rather than the abstract goal of maximizing failures, MUCOREST effectively discovers critical code areas and diverse API behaviors. The experimental results on a benchmark of 10 services show that MUCOREST significantly outperforms state-of-the-art API testing approaches by 11.6-261.1% in the number of discovered API bugs. MUCOREST can generate much fewer API calls to discover the same number of bugs compared to the other approaches. Furthermore, 12.17%-64.09% of the bugs discovered by the other techniques can also be found by MUCOREST., Comment: Accepted by KSE'24