October 2025 arXiv papers — page 185

Tianyu Fan, Xinyao Niu, Yuxiang Zheng, Fengji Zhang

DeepResearch agents represent a transformative AI paradigm, conducting expert-level research through sophisticated reasoning and multi-tool integration. However, evaluating these systems remains critically challenging due to open-ended research scenarios and existing benchmarks that focus on isolated capabilities rather than holistic performance. Unlike trad

Deeparnab Chakrabarty, Jonathan Conroy, Ankita Sarkar

We introduce the aggregated clustering problem, where one is given $T$ instances of a center-based clustering task over the same $n$ points, but under different metrics. The goal is to open $k$ centers to minimize an aggregate of the clustering costs -- e.g., the average or maximum -- where the cost is measured via $k$-center/median/means objectives. More ge

Vinayak Sharma, Ashish Padhy, Lord Sen, Vijay Jagdish Karanjkar

Kolmogorov-Arnold Networks or KANs have shown the ability to outperform classical Deep Neural Networks, while using far fewer trainable parameters for regression problems on scientific domains. Even more powerful has been their interpretability due to their structure being composed of univariate B-Spline functions. This enables us to derive closed-form equat

Bruno Cavalar, Boyang Chen, Andrea Coladangelo, Matthew Gray

We give a meta-complexity characterization of EFI pairs, which are considered the "minimal" primitive in quantum cryptography (and are equivalent to quantum commitments). More precisely, we show that the existence of EFI pairs is equivalent to the following: there exists a non-uniformly samplable distribution over pure states such that the problem of estimat

Zhiqing Cui, Binwu Wang, Qingxiang Liu, Yeqiang Wang

Large language models (LLM) have emerged as a promising avenue for time series forecasting, offering the potential to integrate multimodal data. However, existing LLM-based approaches face notable limitations-such as marginalized role in model architectures, reliance on coarse statistical text prompts, and lack of interpretability. In this work, we introduce

Alexandru Chirvasitu

For an $(n\ge 2)$-dimensional real Banach space $E$ with unit ball $E_{\le 1}$ and a topological space $X$ arbitrary elements in $C(X,E_{\le 1})$ are always expressible as linear combinations of at most three functions valued in the unit sphere $\partial E_{\le 1}$. On the other hand, for normal $X$, $C(X,E_{\le 1})$ can only be the convex hull of $C(X,\part

Haochen Yu, Qiankun Liu, Hongyuan Liu, Jianfei Jiang

Feed-forward paradigms for 3D reconstruction have become a focus of recent research, which learn implicit, fixed view transformations to generate a single scene representation. However, their application to complex driving scenes reveals significant limitations. Two core challenges are responsible for this performance gap. First, the reliance on a fixed view

Askar Ali M, Sachindranath Jayaraman, Himadri Mukherjee

The purpose of this article is to investigate triangularization and simultaneous triangularization of matrices over max algebras using graph theoretic methods. We establish a connection between commutators and commutants with simultaneous triangularization over max algebras. We also define the notion of characteristic polynomial of a collection in terms of t

Šárka Hudecová, Claudia Kirch

We propose a test for a change in the mean for a sequence of functional observations that are only partially observed on subsets of the domain, with no information available on the complement. The framework accommodates important scenarios, including both abrupt and gradual changes. The significance of the test statistic is assessed via a permutation test. I

Thomas Lautenschlager, Nils Friederich, Angelo Jovin Yamachui Sitcheu, Katja Nau

High-throughput toxicity testing offers a fast and cost-effective way to test large amounts of compounds. A key component for such systems is the automated evaluation via machine learning models. In this paper, we address critical challenges in this domain and demonstrate how representations learned via self-supervised learning can effectively identify toxic

Shuangyan Deng, Haizhou Peng, Jiachen Xu, Rui Mao

Multimodal Large Language Models (MLLMs) have made substantial progress in recent years. However, their rigorous evaluation within specialized domains like finance is hindered by the absence of datasets characterized by professional-level knowledge intensity, detailed annotations, and advanced reasoning complexity. To address this critical gap, we introduce

Willem Heijltjes

The Functional Machine Calculus (Heijltjes 2022) is a new approach to unifying the imperative and functional programming paradigms. It extends the lambda-calculus, preserving the key features of confluent reduction and typed termination, to embed computational effects, evaluation strategies, and control flow operations. The first instalment modelled sequenti

Radovan Kukobat, Ranko Škrbić, Sanja Jovičić, Milka Matičić

The role of graphene nanowindows in the nanohydroxyapatite bone scaffold preparation is important for the preparation of mechanically robust scaffolds and implementation in bone recovery. Here, we report a graphene-nanohydroxyapatite (G-nHAP) scaffold synthesized by the hydrothermal method, along with its formation mechanism and promising in vivo application

Dingzhou Li, Lei Chang, Ye Tao

Near-space, which covers altitudes from 20 to 100 kilometers, has been receiving more and more attention because of its special strategic value. Airships and high-altitude balloons are two common types of low-speed vehicles that operate in this region. They can be used for jobs like monitoring, communication, and remote sensing, but they need efficient propu

Gustave Cortal, Alain Finkel

Personal narratives are stories authors construct to make meaning of their experiences. Style, the distinctive way authors use language to express themselves, is fundamental to how these narratives convey subjective experiences. Yet there is a lack of a formal framework for systematically analyzing these stylistic choices. We present a novel approach that fo

Pylyp Cherevan

We consider the diagonal paraproduct arising in the nonlinearity $(u\cdot \nabla) u$ for the three-dimensional Navier-Stokes equations. On scale-critical windows and in the range $1/6 < \delta \le 5/8$ we obtain a log-free estimate at the level $L^2_t {\dot H}^{-1}_x$ for the projection $P_{< N^{1-\delta}} \nabla(u_N \otimes v_N)$, consistent with the critic

Liting Li, Yumeng Wang, Yueheng Sun

Graph-level anomaly detection aims to identify anomalous graphs or subgraphs within graph datasets, playing a vital role in various fields such as fraud detection, review classification, and biochemistry. While Graph Neural Networks (GNNs) have made significant progress in this domain, existing methods rely heavily on large amounts of labeled data, which is

Xiaocong Chen, Siyu Wang, Lina Yao

Reinforcement Learning-based recommender systems (RLRS) offer an effective way to handle sequential recommendation tasks but often face difficulties in real-world settings, where user feedback data can be sub-optimal or sparse. In this paper, we introduce MDT4Rec, an offline RLRS framework that builds on the Decision Transformer (DT) to address two major cha

Renaud Leplaideur

We study metastability for symbolic dynamic. We prove that for a global system given by two independent sub-systems linked by a hole, and for a Lipschitz continuous potential, the global equilibrium state converges, as the hole shrinks, to a convex combination of the two independent equilibria in each component. Two kinds of convergence occur, depending on t

Qing Ke, Tianxing Pan, Jin Mao

The production of knowledge has become increasingly a global endeavor. Yet, location related factors, such as local working environment and national policy designs, may continue to affect what kind of science is being pursued. Here we examine the geography of the production of creative science by country, through the lens of novelty and atypicality proposed

Sarbartha Bhattacharya, Tsao-Hsien Chen

Let $G$ be a split connected reductive over a non-archimedean local field $k$. In this paper we give a description of the depth-$r$ Bernstein center of $G(k)$ for rational depths as a limit of depth-$r$ standard parahoric Hecke algebras, extending our previous work in the integral depths case (arXiv:2407.15128). Using this description, we construct maps from

Wenlin Li, Chengsong Zhao, Najmeh Eshaqi-Sani, Zhiyu Jiang

A tenet of contemporary physics is that novel physics beyond the Standard Model lurks at a scale related to the Planck length. The development and validation of a unified framework that merges general relativity and quantum physics is contingent upon the observation of Planck-scale physics. Here, we present a fully quantum model for measuring the nonstationa

Stéphane Crépey, Samuel Drapeau, Mekonnen Tadese

Carbon pricing has become a central pillar of modern climate policy, with carbon taxes and emissions trading systems (ETS) serving as the two dominant approaches. Although economic theory suggests these instruments are equivalent under idealized assumptions, their performance diverges in practice due to real-world market imperfections. A particularly less ex

Edward Y. Chang, Ethan Y. Chang

Large language models can change answers under harmless edits that matter in practice: RAG outputs flip when passages are reordered, fine-tuning erodes invariances learned at pretraining, debate or chain-of-thought prompts take path-dependent routes, and compiler fusion or reordering perturbs logits near decision boundaries. These failures violate intended i

Xiaocong Chen, Siyu Wang, Lina Yao

Reinforcement learning-based recommender systems (RL4RS) have gained attention for their ability to adapt to dynamic user preferences. However, these systems face challenges, particularly in offline settings, where data inefficiency and reliance on pre-collected trajectories limit their broader applicability. While offline reinforcement learning methods leve

Jingyu Peng, Maolin Wang, Hengyi Cai, Yuchen Li

Small language models (SLMs) are crucial for applications with strict latency and computational constraints, yet achieving high performance remains challenging. Knowledge distillation (KD) can transfer capabilities from large teacher models, but existing methods face a dilemma: off-policy distillation provides high-quality supervision but suffers from exposu

Pylyp Cherevan

We investigate the contribution of the full nonlinearity outside the narrow diagonal zone in the three-dimensional Navier-Stokes equations. We consider the off-diagonal components, including lh, hl, as well as part of the resonant block hh -> l for |xi + eta| >= N^(1-delta). The proof relies on three main elements: (i) six-fold integration by parts in the ph

Emre Can Acikgoz, Cheng Qian, Heng Ji, Dilek Hakkani-Tür

One paradigm of language model (LM) fine-tuning relies on creating large training datasets, under the assumption that high quantity and diversity will enable models to generalize to novel tasks after post-training. In practice, gathering large sets of data is inefficient, and training on them is prohibitively expensive; worse, there is no guarantee that the

Ji Yu, Yang shuo, Xu Yuetonghui, Liu Mengmei

Most current music source separation (MSS) methods rely on supervised learning, limited by training data quantity and quality. Though web-crawling can bring abundant data, platform-level track labeling often causes metadata mismatches, impeding accurate &#34;audio-label&#34; pair acquisition. To address this, we present ACMID: a dataset for MSS generated thr

Yijie Gao, Houqiang Zhong, Tianchi Zhu, Zhengxue Cheng

The demand for semantically rich 3D models of indoor scenes is rapidly growing, driven by applications in augmented reality, virtual reality, and robotics. However, creating them from sparse views remains a challenge due to geometric ambiguity. Existing methods often treat semantics as a passive feature painted on an already-formed, and potentially flawed, g

Guan-Ting Lin, Shih-Yun Shan Kuan, Jiatong Shi, Kai-Wei Chang

While full-duplex speech agents enable natural, low-latency interaction by speaking and listening simultaneously, their consistency and task performance in multi-turn settings remain underexplored. We introduce Full-Duplex-Bench-v2 (FDB-v2), a streaming framework that integrates with an automated examiner that enforces staged goals under two pacing setups (F

Harsh Kavediya, Vighnesh Nayak, Bheeshm Sharma, Balamurugan Palaniappan

Sign language to spoken language audio translation is important to connect the hearing- and speech-challenged humans with others. We consider sign language videos with isolated sign sequences rather than continuous grammatical signing. Such videos are useful in educational applications and sign prompt interfaces. Towards this, we propose IsoSignVid2Aud, a no

Hao Wu, Mao-Jun Yan, Chun-Sheng An, Cheng-Rong Deng

The low-lying light tetraquark states are investigated in the non-relativistic quark model (NRQM) including the pseudoscalar meson exchange, where two different confinement potential schemes, the Cornell potential and the linear potential, are employed, along with the instanton-induced interaction serving as the residual spin-dependent interaction. The numer

Weisen Jiang, Sinno Jialin Pan

This paper introduces MetaDefense, a novel framework for defending against finetuning-based jailbreak attacks in large language models (LLMs). We observe that existing defense mechanisms fail to generalize to harmful queries disguised by unseen attack templates, despite LLMs being capable of distinguishing disguised harmful queries in the embedding space. Ba

Lingjun Liu, Feiran Qin, Owolabi Legunsen, Marcelo d'Amorim

Bugs in compilers, which are critical infrastructure today, can have outsized negative impacts. Mutational fuzzers aid compiler bug detection by systematically mutating compiler inputs, i.e., programs. Their effectiveness depends on the quality of the mutators used. Yet, no prior work used compiler bug histories as a source of mutators. We propose IssueMut,

Santatra Hagamalala Bernardin, Riad Mokadem, Franck Morvan, Hasinarivo Ramanana

Multi-cloud computing systems face significant challenges in ensuring acceptable performance while adhering to tenant budget requirements. This paper proposes a tenant budget-aware (tenant-centric) data replication framework for Multi-Cloud Computing (TCDRM). The proposed strategy dynamically creates data replicas based on predefined thresholds for response

Yuta Shikuri, Hironori Fujisawa

Spatial prediction refers to the estimation of unobserved values from spatially distributed observations. Although recent advances have improved the capacity to model diverse observation types, adoption in practice remains limited in industries that demand interpretability. To mitigate this gap, surrogate models that explain black-box predictors provide a pr

Nilesh Pandey, Dipanjan Basu, Leonard F Register, Sanjay K Banerjee

We present a three-dimensional simulation study of silicon nanowire double quantum dots (DQDs) with leads at T = 2 K, which extends beyond traditional effective mass or quasi-1D and quasi-2D approaches typically applied to bulk or planar geometries. A 3-D Poisson solver is self-consistently coupled to 2-D Schrodinger along slices normal to transport (width *

Houqiang Zhong, Zhenglong Wu, Sihua Fu, Zihan Zheng

3D Gaussian Splatting (3DGS) has recently enabled real-time photorealistic rendering in compact scenes, but scaling to large urban environments introduces severe aliasing artifacts and optimization instability, especially under high-resolution (e.g., 4K) rendering. These artifacts, manifesting as flickering textures and jagged edges, arise from the mismatch

Cathal Doyle

In the Generative Age, the nature of knowledge work is transforming. Traditional models that emphasise the organisation and retrieval of pre-existing information are increasingly inadequate in the face of generative AI (GenAI) systems capable of autonomous content creation. This paper introduces the Knowledge Sculptor (KS), a new professional archetype for H

Kaen Kogashi, Anoop Cherian, Meng-Yu Jennifer Kuo

Real-world scenes often feature multiple humans interacting with multiple objects in ways that are causal, goal-oriented, or cooperative. Yet existing 3D human-object interaction (HOI) benchmarks consider only a fraction of these complex interactions. To close this gap, we present MMHOI -- a large-scale, Multi-human Multi-object Interaction dataset consistin

Joseph M. Carlson, Nitish V. Deshpande, Miguel Rodrigo Castellanos, Robert W. Heath

Dynamic metasurface antennas (DMA) provide low-power beamforming through reconfigurable radiative slots. Each slot has a tunable component that consumes low power compared to typical analog components like phase shifters. This makes DMAs a potential candidate to minimize the power consumption of multiple-input multiple-output (MIMO) antenna arrays. In this p

Sirui Yu, Junrong Wang, Huajin Zhao, Hong Mao

We investigate the quantum transport in a hybrid ring system consisting of a quantum dot (QD) coupled to two Majorana bound states (MBSs) hosted in a topological superconducting nanowire, threaded by a magnetic flux. Utilizing the dissipaton equation-of-motion approach, we demonstrate that the differential conductance shows periodic behavior and its periodic

Yuping Zhou, Siqi Lai, Jindong Han, Hao Liu

The rise of Internet of Vehicles (IoV) technologies is transforming traffic management from isolated control to a collective, multi-vehicle process. At the heart of this shift is multi-vehicle dynamic navigation, which requires simultaneously routing large fleets under evolving traffic conditions. Existing path search algorithms and reinforcement learning me

Luis Gabriel Gallegos Mariñez, Lizardo Valencia Palomo, Luis Cedillo Barrera

Correct description of quarkonia production and kinematics are still one of the most challenging assignments for Quantum Chromodynamics. This document presents a study of the $\Upsilon$(1S), (2S) and (3S) mean transverse momentum ($\langle p_{\mathrm{T}}^{\Upsilon} \rangle$) as a function of the charged particle multiplicity ($N_{\mathrm{Track}}$) in proton-

Shohei Enomoto

Visual prompting (VP) has emerged as a promising parameter-efficient fine-tuning approach for adapting pre-trained vision models to downstream tasks without modifying model parameters. Despite offering advantages like negligible computational overhead and compatibility with black-box models, conventional VP methods typically achieve lower accuracy than other

Anu Agarwal, Mihir Pamnani, Dilek Hakkani-Tur

The undesired memorization of sensitive information by Large Language Models (LLMs) has emphasized the need for safety mechanisms that can regulate model behavior. This has led to the development of machine unlearning techniques that enable models to precisely forget sensitive and unwanted information. For machine unlearning, first-order and second-order opt

Raisa M. Simoes, Timoteo Kelly, Eduardo J. Simoes, Praveen Rao

This paper aims to explore two competing data science methodologies to attempt answering the question, "Which issues contributed most to voters' choice in the 2024 presidential election?" The methodologies involve novel empirical evidence driven by artificial intelligence (AI) techniques. By using two distinct methods based on natural language processing and

Eric Hanchen Jiang, Weixuan Ou, Run Liu, Shengyuan Pang

Safety alignment of large language models currently faces a central challenge: existing alignment techniques often prioritize mitigating responses to harmful prompts at the expense of overcautious behavior, leading models to incorrectly refuse benign requests. A key goal of safe alignment is therefore to improve safety while simultaneously minimizing false r

Tracy Chin, Daniel Qin

We study the class of Lorentzian symmetric polynomials and Lorentzian symmetric functions, which are defined to be symmetric functions for which every truncation of variables is Lorentzian. Similar to the space of Lorentzian polynomials, we show that the space of Lorentzian symmetric polynomials is homeomorphic to a closed Euclidean ball. Our main result is

Suying Liu, Xiaodi Wu, Murphy Yuezhen Niu

Hamiltonian learning is a cornerstone for advancing accurate many-body simulations, improving quantum device performance, and enabling quantum-enhanced sensing. Existing readily deployable quantum metrology techniques primarily focus on achieving Heisenberg-limited precision in one- or two-qubit systems. In contrast, general Hamiltonian learning theories add

Jianghao Lin, Rong Shan, Jiachen Zhu, Yunjia Xi

Survey papers are foundational to the scholarly progress of research communities, offering structured overviews that guide both novices and experts across disciplines. However, the recent surge of AI-generated surveys, especially enabled by large language models (LLMs), has transformed this traditionally labor-intensive genre into a low-effort, high-volume o

Kanglin Ning, Ruzhao Chen, Penghong Wang, Xingtao Wang

Explicitly modeling room background depth as a geometric constraint has proven effective for panoramic depth estimation. However, reconstructing this background depth for regular enclosed regions in a complex indoor scene without external measurements remains an open challenge. To address this, we propose a pose-aware and geometry-constrained framework for p

Hui Zhang, Alex Akridge, Charles J. Horowitz, Jinfeng Liao

Understanding how protons and neutrons are located differently in an atomic nucleus can provide fundamental information on nuclear structure and have far-reaching implications for astrophysics. A precise determination of this important difference, often quantified by the so-called neutron skin thickness, is challenging both theoretically and experimentally.

Zeyu Sun, Jingjing Liang, Weiyi Wang, Chenyao Suo

MLIR (Multi-Level Intermediate Representation) has rapidly become a foundational technology for modern compiler frameworks, enabling extensibility across diverse domains. However, ensuring the correctness and robustness of MLIR itself remains challenging. Existing fuzzing approaches-based on manually crafted templates or rule-based mutations-struggle to gene

Javad Sayyadi, Mahdi Nangir, Mahmood Mohassel Feghhi, Hamid Sayyadi

In this manuscript, we explore the application of model-free reinforcement learning in optimizing secure multiparty computation (SMPC) protocols. SMPC is a crucial tool for performing computations on private data without the need to disclose it, holding significant importance in various domains, including information security and privacy. However, the effici

Valerio La Gatta, Dolev Mutzari, Sarit Kraus, VS Subrahmanian

Adversarial environments require agents to navigate a key strategic trade-off: acquiring information enhances situational awareness, but may simultaneously expose them to threats. To investigate this tension, we formulate a PursuitEvasion-Exposure-Concealment Game (PEEC) in which a pursuer agent must decide when to communicate in order to obtain the evader's

Yuxin Huang, Simeng Wu, Ran Song, Yan Xiang

Generative Information Retrieval is an emerging retrieval paradigm that exhibits remarkable performance in monolingual scenarios.However, applying these methods to multilingual retrieval still encounters two primary challenges, cross-lingual identifier misalignment and identifier inflation. To address these limitations, we propose Multilingual Generative Ret

Aryan Poduri

We present an adaptive scheduler for a single differencing engine (SmartDiff) with two execution modes: (i) in-memory threads and (ii) Dask based parallelism. The scheduler continuously tunes batch size and worker/thread count within fixed CPU and memory budgets to minimize p95 latency. A lightweight preflight profiler estimates bytes/row and I/O rate; an on

Luu Tu Nguyen, Vu Tram Anh Khuong, Thi Bich Phuong Man, Thi Duyen Ngo

Facial micro-expressions, characterized by their subtle and brief nature, are valuable indicators of genuine emotions. Despite their significance in psychology, security, and behavioral analysis, micro-expression recognition remains challenging due to the difficulty of capturing subtle facial movements. Optical flow has been widely employed as an input modal

Renhua Ding, Xiao Yang, Zhengwei Fang, Jun Luo

Large Vision-Language Models (LVLMs) empower autonomous mobile agents, yet their security under realistic mobile deployment constraints remains underexplored. While agents are vulnerable to visual prompt injections, stealthily executing such attacks without requiring system-level privileges remains challenging, as existing methods rely on persistent visual m

Daniel Grier, Daniel M. Kane, Jackson Morris, Anthony Ostuni

We construct a family of distributions $\{\mathcal{D}_n\}_n$ with $\mathcal{D}_n$ over $\{0, 1\}^n$ and a family of depth-$7$ quantum circuits $\{C_n\}_n$ such that $\mathcal{D}_n$ is produced exactly by $C_n$ with the all zeros state as input, yet any constant-depth classical circuit with bounded fan-in gates evaluated on any binary product distribution has

Grace Cai, Nithin Parepally, Laura Zheng, Ming C. Lin

Modeling the dynamics of micro-mobility vehicles (MMV) is becoming increasingly important for training autonomous vehicle systems and building urban traffic simulations. However, mainstream tools rely on variants of the Kinematic Bicycle Model (KBM) or mode-specific physics that miss tire slip, load transfer, and rider/vehicle lean. To our knowledge, no unif

Yihao Peng, Biao Ma, Hai Wan, Xibin Zhao

Modern web application recovery presents a critical dilemma. Coarse-grained snapshot rollbacks cause unacceptable data loss for legitimate users. Surgically removing an attack's impact is hindered by a fundamental challenge in high-concurrency environments: it is difficult to attribute resulting file and database modifications to a specific attack-related re

Zhiyuan Ding

We give an elementary proof of the reducedness of twisted loop groups along the lines of the Kneser-Tits problem.

Bowen Ouyang, Pratik Rath

Previous work on Jackiw-Teitelboim (JT) gravity has shown that, at low temperatures, the annealed entropy becomes negative and departs from the quenched entropy. From the perspective of the random-matrix theory (RMT) dual of JT gravity, this effect is encoded in the continuous spectrum at the spectral edge that is universally described by the Airy model. At

Terence R. Smith

There is consensus that sums $S_n={ {\Sigma }_{k=1}^n R_{0k} e^{i \theta_k}}$ of complex exponential terms, despite their mathematical significance, only possess closed-form representations for specific values of n and special values of their parameters and that there are no generally-accepted recursive formulae for their computation. This note is focused on

Jixing Zhang, Bo Peng, Yang Wang, Cheuk Kit Cheung

As quantum control approaches hardware-imposed performance limits, weak effects omitted by reduced models become consequential. Assumptions required for analytic tractability then cease to guide control design and instead constrain further improvement. Here, we relax such assumptions and use simulation-guided stochastic tree search to navigate combinatoriall

Jinho Cha, Youngchul Kim, Junyeol Ryu, Sangjun Park

This study develops a strategic procurement framework integrating blockchain-based smart contracts with bounded demand variability modeled through a truncated normal distribution. While existing research emphasizes the technical feasibility of smart contracts, the operational and economic implications of adoption under moderate uncertainty remain underexplor

Y. F. Liang, L. T. Yang, Q. Yue, K. J. Kang

We present limits on spin-independent inelastic weakly interacting massive particles (WIMP)-nucleus scattering using the 737.1 kg$\cdot$day dataset from the CDEX-1B experiment. Expected nuclear recoil spectra for various inelastic WIMP masses $m_\chi$ and mass splittings $\delta$ are calculated under the standard halo model. An accurate background model of C

Majid Jaberi-Douraki, Hossein Sholehrasa, Xuan Xu, Remya Ampadi Ramachandran

The extraction and standardization of pharmacokinetic (PK) information from scientific literature remain significant challenges in computational pharmacology, which limits the reliability of data-driven models in drug development. Large language models (LLMs) have achieved remarkable progress in text understanding and reasoning, yet their adaptation to struc

Peilin Wu, Mian Zhang, Kun Wan, Wentian Zhao

Agentic RAG is a powerful technique for incorporating external information that LLMs lack, enabling better problem solving and question answering. However, suboptimal search behaviors exist widely, such as over-search (retrieving information already known) and under-search (failing to search when necessary), which leads to unnecessary overhead and unreliable

Sajib Acharjee Dip, Adrika Zafor, Bikash Kumar Paul, Uddip Acharjee Shuvo

Large language models (LLMs) and emerging agentic frameworks are beginning to transform single-cell biology by enabling natural-language reasoning, generative annotation, and multimodal data integration. However, progress remains fragmented across data modalities, architectures, and evaluation standards. LLM4Cell presents the first unified survey of 58 found

Qinghongbing Xie, Zhaoyuan Xia, Feng Zhu, Lijun Gong

Recently spatial-temporal intelligence of Visual-Language Models (VLMs) has attracted much attention due to its importance for autonomous driving, embodied AI and general AI. Existing spatial-temporal benchmarks mainly focus on egocentric (first-person) perspective reasoning using images/video contexts, or geographic reasoning with graphical context (e.g., m

Hao Wu, Wei Liu

Reinforcement learning has been widely applied to enhance the reasoning capabilities of large language models. Extending the inference limits of smaller models has become a prominent research focus. However, algorithms such as Group Relative Policy Optimization (GRPO) suffer from a clear drawback: the upper bound of a model's rollout responses is entirely de

Angelos Pelecanos, Jack Spilecki, John Wright

In the problem of quantum state tomography, one is given $n$ copies of an unknown rank-$r$ mixed state $\rho \in \mathbb{C}^{d \times d}$ and asked to produce an estimator of $\rho$. In this work, we present the debiased Keyl's algorithm, the first estimator for full state tomography which is both unbiased and sample-optimal. We derive an explicit formula fo

Noy Soffer Aranov

Given a subset $\mathcal{S}\subseteq \mathbb{F}_q[x]$ and fixed integers $n,m\in \mathbb{N}$, we study the distribution of the smallest denominator $Q\in \mathcal{S}$ for which there exists $\mathbf{P}\in \mathbb{F}_q[x]^m$ such that $\left\Vert\frac{\mathbf{P}}{Q}-\boldsymbol{\alpha}\right\Vert<q^{-n}$, where $\boldsymbol{\alpha}\in x^{-1}\mathbb{F}_q((x^{-

Seth Minor, Bret D. Elderd, Benjamin Van Allen, David M. Bortz

Insect species subject to infection, predation, and anisotropic environmental conditions may exhibit preferential movement patterns. Given the innate stochasticity of exogenous factors driving these patterns over short timescales, individual insect trajectories typically obey overdamped stochastic dynamics. In practice, data-driven modeling approaches design

Ming Jie Ong, Sze Yinn Ung, Sim Kuan Goh, Jimmy Y. Zhong

The current study investigated the use of Explainable Artificial Intelligence (XAI) to improve the accuracy of brain tumor segmentation in MRI images, with the goal of assisting physicians in clinical decision-making. The study focused on applying UNet models for brain tumor segmentation and using the XAI techniques of Gradient-weighted Class Activation Mapp

Ruining He, Lukasz Heldt, Lichan Hong, Raghunandan Keshavan

Large Language Models (LLMs) pose a new paradigm of modeling and computation for information tasks. Recommendation systems are a critical application domain poised to benefit significantly from the sequence modeling capabilities and world knowledge inherent in these large models. In this paper, we introduce PLUM, a framework designed to adapt pre-trained LLM

Menglong Zhang, Gennian Ge

A $K_4$-decomposition of a graph is a partition of its edges into $K_4$s. A fractional $K_4$-decomposition is an assignment of a nonnegative weight to each $K_4$ in a graph such that the sum of the weights of the $K_4$s containing any given edge is one. Formulating a nonlinear programming and reducing the number of variables slowly, we prove that every graph

Shuichiro Haruta, Kazunori Matsumoto, Zhi Li, Yanan Wang

In this paper, we propose a rotation-constrained compensation method to address the errors introduced by structured pruning of large language models (LLMs). LLMs are trained on massive datasets and accumulate rich semantic knowledge in their representation space. In contrast, pruning is typically carried out with only a small amount of calibration data, whic

John S. Wallis, David R. Gozzard, Alex M. Frost, Benjamin P. Dix-Matthews

We present a technique for binary spatial mode demultiplexing, using a double-clad fiber coupler as an optical mode sorter, for hypothesis testing for one or two point sources in an incident optical field. By directly coupling an optical field through a double-clad fiber coupler, we demultiplex the field into the fundamental mode and a superposition of highe

Vibhuti Arora, Vinayak M

This paper introduces the second Bohr radius for vector-valued holomorphic functions defined on arbitrary complete Reinhardt domains. We aim to establish the lower and upper bounds of the second Bohr radius in both finite and infinite-dimensional settings. Additionally, we provide specific estimates that connect the Second Bohr radius to a symmetric Banach s

Futoshi Hayasaka, Vijay Kodiyalam

We give lower and upper bounds on the Buchsbaum-Rim multiplicity of finitely generated torsion-free modules over two-dimensional regular local rings, and conditions for them to attain the bounds. As consequences, we have formulae on the multiplicity of integrally closed modules.

Yandu Chen, Kefan Gu, Yuqing Wen, Yucheng Zhao

Vision-Language-Action (VLA) models leverage pretrained vision-language models (VLMs) to couple perception with robotic control, offering a promising path toward general-purpose embodied intelligence. However, current SOTA VLAs are primarily pretrained on multimodal tasks with limited relevance to embodied scenarios, and then finetuned to map explicit instru

Vardhan Dongre, Ryan A. Rossi, Viet Dac Lai, David Seunghyun Yoon

Large Language Models (LLMs) excel at single-turn tasks such as instruction following and summarization, yet real-world deployments require sustained multi-turn interactions where user goals and conversational context persist and evolve. A recurring challenge in this setting is context drift: the gradual divergence of a model's outputs from goal-consistent b

Shiman Zhao, Shangyuan Li, Wei Chen, Tengjiao Wang

Few-shot Multi-label Intent Detection (MID) is crucial for dialogue systems, aiming to detect multiple intents of utterances in low-resource dialogue domains. Previous studies focus on a two-stage pipeline. They first learn representations of utterances with multiple labels and then use a threshold-based strategy to identify multi-label results. However, the

Ishir Rao

Electroencephalography (EEG) is widely used to study human brain dynamics, yet its quantitative information capacity remains unclear. Here, we combine information theory and synthetic forward modeling to estimate the mutual information between latent cortical sources and EEG recordings. Using Gaussian-channel theory and empirical simulations, we find that sc

Omar Mahmoud, Ali Khalil, Buddhika Laknath Semage, Thommen George Karimpanal

Hallucination in large language models (LLMs) has been widely studied in recent years, with progress in both detection and mitigation aimed at improving truthfulness. Yet, a critical side effect remains largely overlooked: enhancing truthfulness can negatively impact safety alignment. In this paper, we investigate this trade-off and show that increasing fact

Youliang Yuan, Qiuyang Mang, Jingbang Chen, Hong Wan

In this paper, we observe that current models are susceptible to reward hacking, leading to a substantial overestimation of a model's reasoning ability. This is evidenced by a high incidence of false positives-solutions that reach the correct answer through an unsound process. Through a systematic analysis with human verification, we establish a taxonomy of

YuHang Tang, Yixuan Lou, Pengfei Han, Haoming Song

Learning manipulation skills from human demonstration videos presents a promising yet challenging problem, primarily due to the significant embodiment gap between human body and robot manipulators. Existing methods rely on paired datasets or hand-crafted rewards, which limit scalability and generalization. We propose TrajSkill, a framework for Trajectory Con

Tianle Zhou, Jiakai Xu, Guanhong Liu, Jiaxiang Liu

Large Language Models (LLMs) suffer from reliability issues on complex tasks, as existing decomposition methods are heuristic and rely on agent or manual decomposition. This work introduces a novel, systematic decomposition framework that we call Analysis of CONstraint-Induced Complexity (ACONIC), which models the task as a constraint problem and leverages f

Bingqiao Gu, Jiale Zeng, Xingqin Qi, Dong Li

Due to the advantages of hypergraphs in modeling high-order relationships in complex systems, they have been applied to higher-order clustering, hypergraph neural networks and computer vision. These applications rely heavily on access to high-quality, large-scale real-world hypergraph data. Yet, compared to traditional pairwise graphs, real hypergraph datase

Long Zhang, Guoying Gao

Altermagnets have recently garnered significant interest due to their vanishing net magnetic moment and non-relativistic momentum-dependent spin splitting. However, altermagnetic (AM) multiferroics especially triferroics remain scarce. We investigate the experimentally synthesized non-van der Waals CrSb as a model system to explore the effects of dimensional

Kathrin Bringmann, Ben Kane, Anubhab Pahari, Larry Rolen

In recent years, there has been extensive work on inequalities among partition functions. In particular, Nicolas, and independently DeSalvo--Pak, proved that the partition function $p(n)$ is eventually log-concave. Inspired by this and other results, Chern--Fu--Tang first conjectured log-concavity of $k$-coloured partitions. Three of the authors and Tripp la

Murong Yue, Zhiwei Liu, Liangwei Yang, Jianguo Zhang

Large Language Models (LLMs) equipped with external tools have demonstrated enhanced performance on complex reasoning tasks. The widespread adoption of this tool-augmented reasoning is hindered by the scarcity of domain-specific tools. For instance, in domains such as physics question answering, suitable and specialized tools are often missing. Recent work h

G. A. Mantashian, D. B. Hayrapetyan, P. A. Mantashyan

Precisely addressing single nanostructures inside dense ensembles remains a bottleneck for scalable photonic and quantum information devices. Here we demonstrate, through comprehensive finite element and variational Monte-Carlo modelling, that a reconfigurable three-dimensional array of needle shaped beams can selectively switch the quantum optical response

Linping Qu, Shenghui Song, Chi-Ying Tsui

In wireless federated learning (FL), the clients need to transmit the high-dimensional deep neural network (DNN) parameters through bandwidth-limited channels, which causes the communication latency issue. In this paper, we propose a layer-wise adaptive modulation scheme to save the communication latency. Unlike existing works which assign the same modulatio

Tianzhe Zhou, Chun Huang

Pulse-profile modeling (PPM) of thermal X-ray emission from rotation-powered millisecond pulsars enables simultaneous constraints on the mass $M$, radius $R$, and hence the equation of state of cold, dense matter. However, Bayesian PPM has faced a hard accuracy-speed bottleneck: current production resolutions used to keep inference tractable can under-resolv