May 2025 arXiv papers — page 60

Yaoning Yu, Ye Yu, Peiyan Zhang, Kai Wei

Prompt quality plays a critical role in the performance of large language models (LLMs), motivating a growing body of work on prompt optimization. Most existing methods optimize prompts over a fixed dataset, assuming static input distributions and offering limited support for iterative improvement. We introduce SIPDO (Self-Improving Prompts through Data-Augm

Yang Zou, Meili Qi, Jianhua Zhang, Difei Zhang

X-ray computed tomography (CT) is widely utilized in the medical, industrial, and other fields to nondestructively generate three-dimensional structural images of objects. However, CT images are often affected by various artifacts, with ring artifacts being a common occurrence that significantly compromises image quality and subsequent structural interpretat

Maitham F. AL-Sunni, Hassan Almubarak, Katherine Horng, John M. Dolan

We present Look-Back and Look-Ahead Adaptive Model Predictive Control (LLA-MPC), a real-time adaptive control framework for autonomous racing that addresses the challenge of rapidly changing tire-surface interactions. Unlike existing approaches requiring substantial data collection or offline training, LLA-MPC employs a model bank for immediate adaptation wi

Aggrey Muhebwa, Khalid K. Osman

Large proprietary language models exhibit strong causal reasoning abilities that smaller open-source models struggle to replicate. We introduce a novel framework for distilling causal explanations that transfers causal reasoning skills from a powerful teacher model to a compact open-source model. The key idea is to train the smaller model to develop causal r

Dongil Yang, Minjin Kim, Sunghwan Kim, Beong-woo Kwak

The remarkable reasoning and generalization capabilities of Large Language Models (LLMs) have paved the way for their expanding applications in embodied AI, robotics, and other real-world tasks. To effectively support these applications, grounding in spatial and temporal understanding in multimodal environments is essential. To this end, recent works have le

Yifan Jia, Kailin Jiang, Yuyang Liang, Qihan Ren

Large Multimodal Models(LMMs) face notable challenges when encountering multimodal knowledge conflicts, particularly under retrieval-augmented generation(RAG) frameworks where the contextual information from external sources may contradict the model's internal parametric knowledge, leading to unreliable outputs. However, existing benchmarks fail to reflect s

Kai Toyosawa, Zhiyuan Yang

Given weakly exact tracial von Neumann algebras $M_{1}, M_{2}$ with a common injective amalgam $B$, we prove that the amalgamated free product $M_{1}\overline{*}_{B}M_{2}$ is biexact relative to $\{M_{1},M_{2}\}$. In the case where $ M_1 $ and $M_2$ are injective, we further show that $M_{1}\overline{*}_{B}M_{2}$ is biexact relative to the amalgam $B$, and i

Chenyu Lu, Shiliang Sun, Jing Zhao, Nan Zhang

Multimodal machine translation (MMT) seeks to address the challenges posed by linguistic polysemy and ambiguity in translation tasks by incorporating visual information. A key bottleneck in current MMT research is the effective utilization of visual data. Previous approaches have focused on extracting global or region-level image features and using attention

Saurabh Belgaonkar, Deepak Prakash Kumar, Sivakumar Rathinam, Swaroop Darbha

This paper presents an integrated approach for efficient path planning and energy management in hybrid unmanned aerial vehicles (HUAVs) equipped with dual fuel-electric propulsion systems. These HUAVs operate in environments that include noise-restricted zones, referred to as quiet zones, where only electric mode is permitted. We address the problem by param

Yu Xia, Rui Zhong, Hao Gu, Wei Yang

Large Language Models (LLMs) have garnered significant attention in Recommendation Systems (RS) due to their extensive world knowledge and robust reasoning capabilities. However, a critical challenge lies in enabling LLMs to effectively comprehend and extract insights from massive user behaviors. Current approaches that directly leverage LLMs for user intere

Hui Ma, Kai Yang, Man-On Pun

Cellular traffic prediction is of great importance for operators to manage network resources and make decisions. Traffic is highly dynamic and influenced by many exogenous factors, which would lead to the degradation of traffic prediction accuracy. This paper proposes an end-to-end framework with two variants to explicitly characterize the spatiotemporal pat

Pingzhi Li, Zhen Tan, Mohan Zhang, Huaizhi Qu

Large Language Models (LLMs) represent substantial intellectual and economic investments, yet their effectiveness can inadvertently facilitate model imitation via knowledge distillation (KD). In practical scenarios, competitors can distill proprietary LLM capabilities by simply observing publicly accessible outputs, akin to reverse-engineering a complex perf

Sanghyun Kim, Deunsol Jung, Minsu Cho

Recent methods for zero-shot Human-Object Interaction (HOI) detection typically leverage the generalization ability of large Vision-Language Model (VLM), i.e., CLIP, on unseen categories, showing impressive results on various zero-shot settings. However, existing methods struggle to adapt CLIP representations for human-object pairs, as CLIP tends to overlook

Guang Yang, Yu Zhou, Xiang Chen, Wei Zheng

Trustworthy evaluation methods for code snippets play a crucial role in neural code generation. Traditional methods, which either rely on reference solutions or require executable test cases, have inherent limitation in flexibility and scalability. The recent LLM-as-Judge methodology offers a promising alternative by directly evaluating functional consistenc

Ming Yin, Yuanhao Qu, Ling Yang, Le Cong

We investigate how to teach large language models (LLMs) to perform scientific reasoning by leveraging expert discussions as a learning signal. Focusing on the genomics domain, we develop an automated pipeline to extract trainable data and introduce Genome-Bench, a new benchmark constructed from over a decade of scientific forum discussions on genome enginee

Tomer Shoham, Katrina Ligettt

Ratio statistics--such as relative risk and odds ratios--play a central role in hypothesis testing, model evaluation, and decision-making across many areas of machine learning, including causal inference and fairness analysis. However, despite privacy concerns surrounding many datasets and despite increasing adoption of differential privacy, differentially p

Shogo Sato, Masaru Tsuchida, Mariko Yamaguchi, Takuhiro Kaneko

Intrinsic image decomposition (IID) is the task of separating an image into albedo and shade. In real-world scenes, it is difficult to quantitatively assess IID quality due to the unavailability of ground truth. The existing method provides the relative reflection intensities based on human-judged annotations. However, these annotations have challenges in su

T-H. Hubert Chan, Shinuo Ma

We study a unified framework for optimization problems defined on dual-modular instances, where the input comprises a finite ground set $V$ and two set functions: a monotone supermodular reward function $\f$ and a strictly monotone submodular cost function $\g$. This abstraction captures and generalizes classical models in economics and combinatorial optimiz

Nanxing Hu, Xiaoyue Duan, Jinchao Zhang, Guoliang Kang

Large Vision-Language Models (LVLMs) usually generate texts which satisfy context coherence but don't match the visual input. Such a hallucination issue hinders LVLMs' applicability in the real world. The key to solving hallucination in LVLM is to make the text generation rely more on the visual content. Most previous works choose to enhance/adjust the featu

Yiqiao Liao, Farinaz Koushanfar, Parinaz Naghizadeh

We introduce DyCO-GNN, a novel unsupervised learning framework for Dynamic Combinatorial Optimization that requires no training data beyond the problem instance itself. DyCO-GNN leverages structural similarities across time-evolving graph snapshots to accelerate optimization while maintaining solution quality. We evaluate DyCO-GNN on dynamic maximum cut, max

Yiyin Cao, Chuangyin Dang

A weakening of sequential rationality of sequential equilibrium yields Reny's (1992) weakly sequentially rational equilibrium (WSRE) in extensive-form games. WSRE requires Kreps and Wilson's (1982) consistent assessment to satisfy global rationality of nonconvex payoff functions at every information set reachable by a player's own strategy. The consistent as

Hui Ma, Kai Yang

Network traffic prediction techniques have attracted much attention since they are valuable for network congestion control and user experience improvement. While existing prediction techniques can achieve favorable performance when there is sufficient training data, it remains a great challenge to make accurate predictions when only a small amount of trainin

Wei Li, Dezhao Luo, Dongbao Yang, Zhenhang Li

Video action understanding tasks in real-world scenarios always suffer data limitations. In this paper, we address the data-limited action understanding problem by bridging data scarcity. We propose a novel method that employs a text-to-video diffusion transformer to generate annotated data for model training. This paradigm enables the generation of realisti

Manoj Balaji Jagadeeshan, Prince Raj, Pawan Goyal

The study presents a comprehensive benchmark for retrieving Sanskrit documents using English queries, focusing on the chapters of the Srimadbhagavatam. It employs a tripartite approach: Direct Retrieval (DR), Translation-based Retrieval (DT), and Query Translation (QT), utilizing shared embedding spaces and advanced translation methods to enhance retrieval s

Fei Zhao, Xueliang Zhang, Zhong-Qiu Wang

Acoustic echo cancellation (AEC) is an important speech signal processing technology that can remove echoes from microphone signals to enable natural-sounding full-duplex speech communication. While single-channel AEC is widely adopted, multi-channel AEC can leverage spatial cues afforded by multiple microphones to achieve better performance. Existing multi-

Chuang Wang, Haitao Zhou, Ling Luo, Qian Yu

3D vector graphics play a crucial role in various applications including 3D shape retrieval, conceptual design, and virtual reality interactions due to their ability to capture essential structural information with minimal representation. While recent approaches have shown promise in generating 3D vector graphics, they often suffer from lengthy processing ti

Wenhao Yang, Sifan Yang, Lijun Zhang

Reflecting the greater significance of recent history over the distant past in non-stationary environments, $\lambda$-discounted regret has been introduced in online convex optimization (OCO) to gracefully forget past data as new information arrives. When the discount factor $\lambda$ is given, online gradient descent with an appropriate step size achieves a

Jianxing Liao, Junyan Xu, Yatao Sun, Maowen Tang

Designing complex computer-aided design (CAD) models is often time-consuming due to challenges such as computational inefficiency and the difficulty of generating precise models. We propose a novel language-guided framework for industrial design automation to address these issues, integrating large language models (LLMs) with computer-automated design (CAuto

Shu Zhong, Mingyu Xu, Tenglong Ao, Guang Shi

In this paper, we share our reflections and insights on understanding Transformer architectures through the lens of associative memory--a classic psychological concept inspired by human cognition. We start with the basics of associative memory (think simple linear attention) and then dive into two dimensions: Memory Capacity: How much can a Transformer reall

Zhuoheng Gao, Yihao Li, Jiyao Zhang, Rui Zhao

Conventional frame-based cameras often struggle with stereo depth estimation in rapidly changing scenes. In contrast, bio-inspired spike cameras emit asynchronous events at microsecond-level resolution, providing an alternative sensing modality. However, existing methods lack specialized stereo algorithms and benchmarks tailored to the spike data. To address

Maonan Wang, Yirong Chen, Aoyu Pang, Yuxin Cai

Traffic signal control (TSC) is a core challenge in urban mobility, where real-time decisions must balance efficiency and safety. Existing methods - ranging from rule-based heuristics to reinforcement learning (RL) - often struggle to generalize to complex, dynamic, and safety-critical scenarios. We introduce VLMLight, a novel TSC framework that integrates v

Mohammed M. Al Ezzi, Albert Zhu, Daniel Bennett, Daniel T. Larson

Two-dimensional multi-layer materials with an induced moir\'e pattern, either due to strain or relative twist between layers, provide a versatile platform for exploring strongly correlated and topological electronic phenomena. While these systems offer unprecedented tunability, their theoretical description remains challenging due to their complex atomic str

Ruixiang Feng, Shen Gao, Xiuying Chen, Lisi Chen

Large Language Models (LLMs) have demonstrated remarkable capabilities across various tasks, yet they often exhibit a specific cultural biases, neglecting the values and linguistic diversity of low-resource regions. This cultural bias not only undermines universal equality, but also risks reinforcing stereotypes and perpetuating discrimination. To address th

Greta X. Li, Elijah J. Kolmes, Ian E. Ochs, Nathaniel J. Fisch

Steady state distribution functions can be used to calculate stability conditions for modes, radiation energy losses, and particle loss rates. Heuristic analytic approximations to these distributions can capture key behaviors of the true distributions such as the relative speeds of different transport processes while possessing computational advantages over

Di Zhao, Bo Jiang, Song Liu, Susu Cui

The increasing demand for privacy protection and security considerations leads to a significant rise in the proportion of encrypted network traffic. Since traffic content becomes unrecognizable after encryption, accurate analysis is challenging, making it difficult to classify applications and detect attacks. Deep learning is currently the predominant approa

Diogo Cruz

We investigate the look-ahead capabilities of chess-playing neural networks, specifically focusing on the Leela Chess Zero policy network. We build on the work of Jenner et al. (2024) by analyzing the model's ability to consider future moves and alternative sequences beyond the immediate next move. Our findings reveal that the network's look-ahead behavior i

Hao Kang, Qingru Zhang, Han Cai, Weiyuan Xu

Large language models (LLMs) have shown remarkable performance across diverse reasoning and generation tasks, and are increasingly deployed as agents in dynamic environments such as code generation and recommendation systems. However, many real-world applications, such as high-frequency trading and real-time competitive gaming, require decisions under strict

Fei Zhao, Shulin He, Xueliang Zhang

Data-driven acoustic echo cancellation (AEC) methods, predominantly trained on synthetic or constrained real-world datasets, encounter performance declines in unseen echo scenarios, especially in real environments where echo paths are not directly observable. Our proposed method counters this limitation by integrating room impulse response (RIR) as a pivotal

Lakshmi Aishwarya Malladi, Navarun Gupta, Ahmed El-Sayed, Xingguo Xiong

Over 8,024 wildfire incidents have been documented in 2024 alone, affecting thousands of fatalities and significant damage to infrastructure and ecosystems. Wildfires in the United States have inflicted devastating losses. Wildfires are becoming more frequent and intense, which highlights how urgently efficient warning systems are needed to avoid disastrous

Haider A. H. Alobaidy, Mehran Behjati, Rosdiadee Nordin, Muhammad Aidiel Zulkifley

This work proposes an empirical air to ground (A2G) propagation model specifically designed for cellular connected unmanned aerial vehicles (UAVs). An in depth aerial drive test was carried out within an operating Long Term Evolution (LTE) network, gathering thorough measurements of key network parameters. Rigid preprocessing and statistical analysis of thes

Chiyu Ma, Enpei Zhang, Yilun Zhao, Wenjun Liu

LLM-as-Judge has emerged as a scalable alternative to human evaluation, enabling large language models (LLMs) to provide reward signals in trainings. While recent work has explored multi-agent extensions such as multi-agent debate and meta-judging to enhance evaluation quality, the question of how intrinsic biases manifest in these settings remains underexpl

Ziqian Wang, Zikai Liu, Xinfa Zhu, Yike Zhu

Generative models have excelled in audio tasks using approaches such as language models, diffusion, and flow matching. However, existing generative approaches for speech enhancement (SE) face notable challenges: language model-based methods suffer from quantization loss, leading to compromised speaker similarity and intelligibility, while diffusion models re

Mohammad Mahdi Moradi, Hossam Amer, Sudhir Mudur, Weiwei Zhang

Learning to adapt pretrained language models to unlabeled, out-of-distribution data is a critical challenge, as models often falter on structurally novel reasoning tasks even while excelling within their training distribution. We introduce a new framework called VDS-TTT - Verifier-Driven Sample Selection for Test-Time Training to efficiently address this. We

Xinmiao Hu, Chun Wang, Ruihe An, ChenYu Shao

Multimodal Large Language Models (MLLMs) have demonstrated strong performance in visual understanding tasks, yet they often suffer from object hallucinations--generating descriptions of objects that are inconsistent with or entirely absent from the input. This issue is closely related to dataset biases, where frequent co-occurrences of objects lead to entang

Haoran Xin, Ying Sun, Chao Wang, Yanke Yu

Despite the success of recommender systems in alleviating information overload, fairness issues have raised concerns in recent years, potentially leading to unequal treatment for certain user groups. While efforts have been made to improve recommendation fairness, they often assume that users' sensitive attributes are available during model training. However

Mohammad Mahdi Moradi, Walid Ahmed, Shuangyue Wen, Sudhir Mudur

Attention and State-Space Models (SSMs) when combined in a hybrid network in sequence or in parallel provide complementary strengths. In a hybrid sequential pipeline they alternate between applying a transformer to the input and then feeding its output into a SSM. This results in idle periods in the individual components increasing end-to-end latency and low

Alessandra Calin, Ian Cartwright, Luke Coffman, Alonso Delfín

We present a generalization of H\"older duality to algebra-valued pairings via $L^p$-modules. H\"older duality states that if $p \in (1, \infty)$ and $p^{\prime}$ are conjugate exponents, then the dual space of $L^p(\mu)$ is isometrically isomorphic to $L^{p^{\prime}}(\mu)$. In this work we study certain pairs $(\mathsf{Y},\mathsf{X})$, as generalizations of

Futoshi Futami, Masahiro Fujisawa

Latent variables (LVs) play a crucial role in encoder-decoder models by enabling effective data compression, prediction, and generation. Although their theoretical properties, such as generalization, have been extensively studied in supervised learning, similar analyses for unsupervised models such as variational autoencoders (VAEs) remain insufficiently und

Mingzhuo Li, Guang Li, Jiafeng Mao, Takahiro Ogawa

Dataset distillation enables the training of deep neural networks with comparable performance in significantly reduced time by compressing large datasets into small and representative ones. Although the introduction of generative models has made great achievements in this field, the distributions of their distilled datasets are not diverse enough to represen

Masahito Kubo, Daikou Shiota, Yukio Katsukawa, Masumi Shimojo

The detailed investigation of the polar magnetic field and its time evolution is one of the major achievements of Hinode. Precise measurements of the polar magnetic field are essential for understanding the solar cycle, as they provide important constraints for identifying the source regions of the solar wind. The Spectropolarimeter (SP) of the Solar Optical

Jia Yin, Khaled Z. Ibrahim, Mauro Del Ben, Jack Deslippe

The numerical solution of the Kadanoff-Baym nonlinear integro-differential equations, which yields the non-equilibrium Green's functions (NEGFs) of quantum many-body systems, poses significant computational challenges due to its high computational complexity. In this work, we present efficient implementations of a numerical method for solving these equations

Jifeng Hu, Sili Huang, Siyuan Guo, Zhaogeng Liu

In recent years, generative models have shown remarkable capabilities across diverse fields, including images, videos, language, and decision-making. By applying powerful generative models such as flow-based models to reinforcement learning, we can effectively model complex multi-modal action distributions and achieve superior robotic control in continuous a

Hongyu Liang, Yuting Zheng, Yihan Li, Yiran Zhang

As large language models (LLMs) continue to advance, their deployment often involves fine-tuning to enhance performance on specific downstream tasks. However, this customization is sometimes accompanied by misleading claims about the origins, raising significant concerns about transparency and trust within the open-source community. Existing model verificati

Hengwei Zhang, Minghui Wu, Li Qiao, Ling Liu

This letter proposes a deep-learning (DL)-based multi-user channel state information (CSI) feedback framework for massive multiple-input multiple-output systems, where the deep joint source-channel coding (DJSCC) is utilized to improve the CSI reconstruction accuracy. Specifically, we design a multi-user joint CSI feedback framework, whereby the CSI correlat

Haoran Xin, Ying Sun, Chao Wang, Weijia Zhang

Incorporating collaborative information (CI) effectively is crucial for leveraging LLMs in recommendation tasks. Existing approaches often encode CI using soft tokens or abstract identifiers, which introduces a semantic misalignment with the LLM's natural language pretraining and hampers knowledge integration. To address this, we propose expressing CI direct

Haoyu Zhao, Sixu Lin, Qingwei Ben, Minyue Dai

This paper presents a novel framework that enables real-world humanoid robots to maintain stability while performing human-like motion. Current methods train a policy which allows humanoid robots to follow human body using the massive retargeted human data via reinforcement learning. However, due to the heterogeneity between human and humanoid robot motion,

Puyuan Peng, Shang-Wen Li, Abdelrahman Mohamed, David Harwath

We present VoiceStar, the first zero-shot TTS model that achieves both output duration control and extrapolation. VoiceStar is an autoregressive encoder-decoder neural codec language model, that leverages a novel Progress-Monitoring Rotary Position Embedding (PM-RoPE) and is trained with Continuation-Prompt Mixed (CPM) training. PM-RoPE enables the model to

Luming Meng, Boping Liu, Qiong Luo

In the crowded eukaryotic nucleus, euchromatin and heterochromatin segregate into distinct compartments, a phenomenon often attributed to homotypic interactions mediated by liquid liquid phase separation of chromatin associated proteins. Here, we revisit genome compartmentalization by examining the role of in vivo DNA packing density fluctuations driven by A

George Cao

The distinguished weights form a subset of the weight lattice and are closely tied to the notion of $p$-cells. These weights are defined via iterations of the Lusztig-Vogan bijection. We prove that all distinguished weights exhibit an anti-symmetry under the composition of reversal and negation. We show that the distribution of these weights follows a polyno

Kaichao Jiang, He Wang, Xiaoshuai Hao, Xiulong Yang

Joint Energy-based Models (JEMs) are well known for their ability to unify classification and generation within a single framework. Despite their promising generative and discriminative performance, their robustness remains far inferior to adversarial training (AT), which, conversely, achieves strong robustness but sacrifices clean accuracy and lacks generat

Akiyoshi Tomihari, Ryo Karakida

The theoretical understanding of self-attention (SA) has been steadily progressing. A prominent line of work studies a class of SA layers that admit an energy function decreased by state updates. While it provides valuable insights into inherent biases in signal propagation, it often relies on idealized assumptions or additional constraints not necessarily p

Guilong Lu, Xuntao Guo, Rongjunchen Zhang, Wenqiao Zhu

Large language models excel in general tasks, yet assessing their reliability in logic-heavy, precision-critical domains like finance, law, and healthcare remains challenging. To address this, we introduce BizFinBench, the first benchmark specifically designed to evaluate LLMs in real-world financial applications. BizFinBench consists of 6,781 well-annotated

Chong Guan

JavaScript, a scripting language employed to augment the capabilities of web browsers within web pages or browser extensions, utilizes code segments termed JavaScript inclusions. While the security aspects of JavaScript inclusions in web pages have undergone substantial scrutiny, a thorough investigation into the security of such inclusions within browser ex

Xu Li, Fan Lyu

Continual Visual Question Answering (CVQA) based on pre-trained models(PTMs) has achieved promising progress by leveraging prompt tuning to enable continual multi-modal learning. However, most existing methods adopt cross-modal prompt isolation, constructing visual and textual prompts separately, which exacerbates modality imbalance and leads to degraded per

Thomas L. Ahrens, Ian M. Down, Manoranjan Majji

This paper details a methodology to transcribe an optimal control problem into a nonlinear program for generation of the trajectories that optimize a given functional by approximating only the highest order derivatives of a given system's dynamics. The underlying method uses orthogonal polynomial integral collocation by which successive integrals are taken t

Brendan Lucier, Raghuvansh R. Saxena

Two sellers compete to sell identical products to a single buyer. Each seller chooses an arbitrary mechanism, possibly involving lotteries, to sell their product. The utility-maximizing buyer can choose to participate in one or both mechanisms, resolving them in either order. Given a common prior over buyer values, how should the sellers design their mechani

Lida Liu, Jingwei Wang, Yuhao Jing, Songzi Lin

Space group theory is pivotal in the design of nanophotonics devices, enabling the characterization of periodic optical structures such as photonic crystals. The aim of this study is to extend the application of nonsymmorphic space groups in the field of numerical analysis for research and design of nanophotonics devices. In this work, we introduce the nonsy

Shijie Bao, Qi'an Guan, Lin Zhou

In this paper, we generalize Zhou valuations, originally defined on complex domains, to the framework of general schemes. We demonstrate that an algebraic version of the Jonsson--Musta\c{t}\u{a} conjecture is equivalent to the statement that every Zhou valuation is quasi-monomial. By introducing a mixed version of jumping numbers and Tian functions associate

Yun Soo Myung

We obtain quantum parameter ($\alpha$)-mass ($M$) induced spontaneous scalarization of quantum Oppenheimer-Snyder (qOS)-black hole in the Einstein-Gauss-Bonnet-scalar theory with the unknown qOS action. We derive Smarr formula which describes a correct thermodynamics for the bald qOS-black hole. It is turned out that two Davies points of heat capacity are id

A. I. Dubikovsky, P. K. Silaev

We have constructed an approximate analytical solution of the spectral problem for a finite-dimensional matrix of a special kind, which turns out to be a very simple and quite satisfactory model of the metastable state. Most of the characteristic properties of the metastable state are reproduced: line shape, decay dynamics, and density of states. The correct

Hengtong Shen, Haiyan Gu, Haitao Li, Yi Yang

Self-Supervised Learning (SSL) enables us to pre-train foundation models without costly labeled data. Among SSL methods, Contrastive Learning (CL) methods are better at obtaining accurate semantic representations in noise interference. However, due to the significant domain gap, while CL methods have achieved great success in many computer vision tasks, they

Rishabh Bhattacharya, Hari Shankar, Vaishnavi Shivkumar, Ponnurangam Kumaraguru

The growing adoption of Graph Neural Networks (GNNs) in high-stakes domains like healthcare and finance demands reliable explanations of their decision-making processes. While inherently interpretable GNN architectures like Graph Multi-linear Networks (GMT) have emerged, they remain vulnerable to generating explanations based on spurious correlations, potent

Soshun Inoue, Takeshi Inaoka, Hajime Ishihara

Hot carrier generation in metals, where high-energy electron-hole pairs are produced via plasmon excitation, has emerged as a promising mechanism for photoelectric conversion and photocatalysis. However, conventional theories often describe this process through phenomenological relaxation via Landau damping, which fails to account for the microscopic origin

Shriyank Somvanshi, Md Monzurul Islam, Syed Aaqib Javed, Gaurab Chhetri

Bio-inspired algorithms utilize natural processes such as evolution, swarm behavior, foraging, and plant growth to solve complex, nonlinear, high-dimensional optimization problems. However, a plethora of these algorithms require a more rigorous review before making them applicable to the relevant fields. This survey categorizes these algorithms into eight gr

Ranjan Sapkota, Konstantinos I. Roumeliotis, Manoj Karkee

This review presents a comprehensive analysis of two emerging paradigms in AI-assisted software development: vibe coding and agentic coding. While both leverage large language models (LLMs), they differ fundamentally in autonomy, architectural design, and the role of the developer. Vibe coding emphasizes intuitive, human-in-the-loop interaction through promp

Dutao Zhang, Nicolas Rafael Arroyo Arias, YuLong He, Sergey Kovalchuk

Controllable code generation, the ability to synthesize code that follows a specified style while maintaining functionality, remains a challenging task. We propose a two-stage training framework combining contrastive learning and conditional decoding to enable flexible style control. The first stage aligns code style representations with semantic and structu

Jing Nathan Yan, Emma Harvey, Junxiong Wang, Jeffrey M. Rzeszotarski

Recommender systems (RS), which are widely deployed across high-stakes domains, are susceptible to biases that can cause large-scale societal impacts. Researchers have proposed methods to measure and mitigate such biases - but translating academic theory into practice is inherently challenging. Through a semi-structured interview study (N=11), we map the RS

Shashata Sawmya, Micah Adler, Nir Shavit

This paper studies the emergence of interpretable categorical features within large language models (LLMs), analyzing their behavior across training checkpoints (time), transformer layers (space), and varying model sizes (scale). Using sparse autoencoders for mechanistic interpretability, we identify when and where specific semantic concepts emerge within ne

Rihui Xin, Han Liu, Zecheng Wang, Yupeng Zhang

Large Language Models (LLMs) have achieved remarkable success in natural language processing tasks, with Reinforcement Learning (RL) playing a key role in adapting them to specific applications. In mathematical problem solving, however, the reliance on ground truth answers poses significant challenges due to their high collection cost and limited availabilit

Chenguang Wang, Kaiyuan Cui, Weichen Zhao, Tianshu Yu

Sampling from binary quadratic distributions (BQDs) is a fundamental but challenging problem in discrete optimization and probabilistic inference. Previous work established theoretical guarantees for stochastic localization (SL) in continuous domains, where MCMC methods efficiently estimate the required posterior expectations during SL iterations. However, a

Haoqin Sun, Jingguang Tian, Jiaming Zhou, Hui Wang

The Contrastive Language-Audio Pretraining (CLAP) model has demonstrated excellent performance in general audio description-related tasks, such as audio retrieval. However, in the emerging field of emotional speaking style description (ESSD), cross-modal contrastive pretraining remains largely unexplored. In this paper, we propose a novel speech retrieval ta

Ye Ye

Large Language Models (LLMs) falter in multi-step interactions -- often hallucinating, repeating actions, or misinterpreting user corrections -- due to reliance on linear, unstructured context. This fragility stems from the lack of persistent memory to track evolving goals and task dependencies, undermining trust in autonomous agents. We introduce the Task M

X. Feng, D. Zhang, S. Hu, X. Li

Effectively modeling and utilizing spatiotemporal features from RGB and other modalities (\eg, depth, thermal, and event data, denoted as X) is the core of RGB-X tracker design. Existing methods often employ two parallel branches to separately process the RGB and X input streams, requiring the model to simultaneously handle two dispersed feature spaces, whic

Peijie Dong, Zhenheng Tang, Xiang Liu, Lujun Li

Post-training compression reduces the computational and memory costs of large language models (LLMs), enabling resource-efficient deployment. However, existing compression benchmarks only focus on language modeling (e.g., perplexity) and natural language understanding tasks (e.g., GLUE accuracy), ignoring the agentic capabilities - workflow, tool use/functio

Hao Wu, Yuan Gao, Ruijian Gou, Xian Wu

Reliable long-term forecasting of Earth system dynamics is fundamentally limited by instabilities in current artificial intelligence (AI) models during extended autoregressive simulations. These failures often originate from inherent spectral bias, leading to inadequate representation of critical high-frequency, small-scale processes and subsequent uncontrol

Chenguang Wang, Xiaoyu Zhang, Kaiyuan Cui, Weichen Zhao

Training neural samplers directly from unnormalized densities without access to target distribution samples presents a significant challenge. A critical desideratum in these settings is achieving comprehensive mode coverage, ensuring the sampler captures the full diversity of the target distribution. However, prevailing methods often circumvent the lack of t

Keane Ong, Rui Mao, Deeksha Varshney, Paul Pu Liang

Counterfactual reasoning typically involves considering alternatives to actual events. While often applied to understand past events, a distinct form-forward counterfactual reasoning-focuses on anticipating plausible future developments. This type of reasoning is invaluable in dynamic financial markets, where anticipating market developments can powerfully u

Younghan Park, Anuj Diwan, David Harwath, Eunsol Choi

Podcasts have become daily companions for half a billion users. Given the enormous amount of podcast content available, highlights provide a valuable signal that helps viewers get the gist of an episode and decide if they want to invest in listening to it in its entirety. However, identifying highlights automatically is challenging due to the unstructured an

Abhijnan Nath, Carine Graff, Andrei Bachinin, Nikhil Krishnaswamy

AI support of collaborative interactions entails mediating potential misalignment between interlocutor beliefs. Common preference alignment methods like DPO excel in static settings, but struggle in dynamic collaborative tasks where the explicit signals of interlocutor beliefs are sparse and skewed. We propose the Frictional Agent Alignment Framework (FAAF),

Sihan Chen, Dan Zhao, Jongwoo Ko, Colby Banbury

The growing computational demands of large language models (LLMs) make efficient inference and activation strategies increasingly critical. While recent approaches, such as Mixture-of-Experts (MoE), leverage selective activation but require specialized training, training-free sparse activation methods offer broader applicability and superior resource efficie

Wenyang Xiao, Haoyu Zhao, Lingxiao Huang

In-context learning (ICL) is a crucial capability of current large language models (LLMs), where the selection of examples plays a key role in performance. While most existing approaches focus on selecting the most similar examples to the query, the impact of diversity in example selection remains underexplored. We systematically investigate the role of dive

Yuhao He, Jinyu Tian, Haiwei Wu, Jianqing Li

The rapid advancement of diffusion models has enhanced their image inpainting and editing capabilities but also introduced significant societal risks. Adversaries can exploit user images from social media to generate misleading or harmful content. While adversarial perturbations can disrupt inpainting, global perturbation-based methods fail in mask-guided ed

Shimpei Goto

We present a matrix-product-state-based numerical approach for simulating systems composed of several qubits and a common one-dimensional waveguide. In the presented approach, the one-dimensional waveguide is modeled in real space. Thus, one can use the advantage of matrix-product states that are suited for simulating low-entangled one-dimensional systems. T

Bingdong Li, Mei Jiang, Hong Qian, Ke Tang

Evolutionary Reinforcement Learning (ERL), training the Reinforcement Learning (RL) policies with Evolutionary Algorithms (EAs), have demonstrated enhanced exploration capabilities and greater robustness than using traditional policy gradient. However, ERL suffers from the high computational costs and low search efficiency, as EAs require evaluating numerous

Bardia Safaei, Vibashan VS, Vishal M. Patel

Active Domain Adaptation (ADA) adapts models to target domains by selectively labeling a few target samples. Existing ADA methods prioritize uncertain samples but overlook confident ones, which often match ground-truth. We find that incorporating confident predictions into the labeled set before active sampling reduces the search space and improves adaptatio

Wenhua Wu, Chenpeng Su, Siting Zhu, Tianchen Deng

Recent advances in neural radiation fields (NeRF) and 3D Gaussian-based SLAM have achieved impressive localization accuracy and high-quality dense mapping in static scenes. However, these methods remain challenged in dynamic environments, where moving objects violate the static-world assumption and introduce inconsistent observations that degrade both camera

Baichuan Li, Larry Powell, Tracy Hammond

The quality of training data is critical to the performance of machine learning applications in domains like transportation, healthcare, and robotics. Accurate image labeling, however, often relies on time-consuming, expert-driven methods with limited feedback. This research introduces a sketch-based annotation approach supported by large language models (LL

Sudipta Paria, Md Rezoan Ferdous, Aritra Dasgupta, Atri Chatterjee

Scan-based Design-for-Testability (DFT) measures are prevalent in modern digital integrated circuits to achieve high test quality at low hardware cost. With the advent of 3D heterogeneous integration and chiplet-based systems, the role of scan is becoming ever more important due to its ability to make internal design nodes controllable and observable in a sy

Shuqi Fan, Jinzhi Lei

Cell size control is crucial for maintaining cellular function and homeostasis. In this study, we develop a first-order partial differential equation model to examine the effects of three key size control mechanisms: the sizer, timer, and adder. Each mechanism is incorporated into the model through distinct boundary conditions. Exact solutions for these mech

Hang Hua, Ziyun Zeng, Yizhi Song, Yunlong Tang

Recent multimodal image generators such as GPT-4o, Gemini 2.0 Flash, and Gemini 2.5 Pro excel at following complex instructions, editing images and maintaining concept consistency. However, they are still evaluated by disjoint toolkits: text-to-image (T2I) benchmarks that lacks multi-modal conditioning, and customized image generation benchmarks that overloo