arXiv subjects

Meng Xiao

Publications and source records attributed to Meng Xiao.

At least 19 recordsLinked to original sources

arXiv2026

Configurational-space separation and structure selection in three hard squares

Self-assembly of hard particles with diverse shapes gives rise to a rich variety of structures through excluded-volume constraints alone. Here we show that even a minimal system of three hard squares confined in a two-dimensional periodic box exhibits nontrivial configurational behavior relevant to structure selection. As the packing fraction increases, radial distribution functions obtained from Markov-chain Monte Carlo and uniform non-overlapping insertion sampling agree at low densities, deviate markedly over an intermediate range, and converge again at higher densities. Pressure measurements provide strong numerical evidence that the discrepancy originates from the separation of the allowed configurational space into two disconnected regions above a characteristic density. We identify the separation density as $\phi_{\rm sep}=3/5$, construct explicit overlap-free transition pathways connecting the two regions immediately below it, and quantify their relative configurational-space volumes. At higher packing fractions, an approximately L-shaped arrangement of the particle centers becomes strongly favored over a staggered one, revealing a structural motif characteristic of tetratic and square-lattice ordering in larger hard-square systems. These results show that excluded-volume geometry can govern both configurational connectivity and local structure selection even in a three-particle system, revealing how signatures of many-particle self-assembly can already emerge in the few-particle limit.

cond-mat.soft↗

arXiv2026

OntologyAligner: Ontology-Aligned Retrieval and Hierarchy-Guided Large Language Model Reranking for Biomedical Ontology Normalization

Biomedical ontology normalization maps free-text expressions to standardized concepts, enabling consistent integration and analysis of biomedical data. This task remains challenging because lexical variation and subtle distinctions among hierarchically related concepts can obscure concept boundaries. We present OntologyAligner, a three-stage framework that combines ontology-aligned retrieval, large language model candidate reranking, and selective hierarchy-guided refinement. We also construct PhenoNormBench, a unified benchmark comprising 13,390 samples from seven Human Phenotype Ontology datasets. OntologyAligner achieved state-of-the-art performance on HPO normalization, with 88.78% Macro Top-1 Accuracy and 86.75% Micro Top-1 Accuracy, exceeding the strongest baseline by 4.85 and 5.07 percentage points, respectively. Ablation analyses showed complementary contributions from all three stages, and sensitivity analyses demonstrated stability across candidate-set sizes and model backbones. Applications to MONDO, MEDIC, and NCBITaxon further established portability to other ontologies. OntologyAligner offers a generalizable framework for accurate mapping of biomedical text to structured ontology concepts. PhenoNormBench and the code are publicly available at https://github.com/zhelishisongjie/OntologyAligner.

cs.AI↗

arXiv2026

Heralded Free-Electron Writing of the Most Subradiant State in an Atomic Array

The most subradiant eigenstate of a finite subwavelength atomic chain in free space, protected by strongly suppressed radiative decay, offers a powerful resource for photon storage, quantum sensing, and many-body quantum optics. Yet its optical preparation is hindered by the simultaneous need to match a wave vector outside the light cone and a nonuniform envelope. Here, we show that a free electron can overcome these constraints: its velocity sets the imprinted wave vector, while the trajectory of the diffracting wave packet shapes the excitation envelope. This simultaneous momentum and envelope matching enables heralded preparation with near-unity conditional fidelity ($F>99.5\%$) even in a deeply subwavelength regime that is difficult to access with propagating free-space photons. We further show that a path-superposed free electron can excite an antisymmetric state in two closely spaced parallel chains, whose interchain destructive interference yields stronger subradiance than a single chain with the same total number of atoms. These results establish free electrons as quantum writers for collective excitations that are difficult to access with propagating optical fields.

quant-ph↗

arXiv2026

BioHarness: Substrate-Aware Evidence Assembly for Biomedical Question Answering across Literature, Knowledge Bases, and Biological Atlases

Motivation: Biomedical question answering often requires evidence beyond topically retrieved literature, including gene alias resolution, database identifier normalization, and atlas-derived biological measurements. However, existing retrieval-augmented generation (RAG) systems typically follow a fixed workflow and lack an explicit mechanism for deciding when retrieved text is sufficient, when curated biomedical knowledge is required, or when executable evidence assembly over structured measurements should be invoked. This motivates a substrate-aware large language model (LLM) harness that selectively assembles sufficient evidence across literature, knowledge bases, and biological atlases. Results: We introduce BioHarness, an LLM harness for staged biomedical evidence assembly across literature retrieval, curated biomedical knowledge resources, and atlas-derived structured measurements. BioHarness first attempts to answer from reranked literature evidence and escalates through grounded cascade control to REPL-style evidence assembly only when the current evidence is uncertain, weakly grounded, or substrate-mismatched. Across 19,302 biomedical QA items spanning seven answer formats, BioHarness improves the pooled score from 65.9 to 71.0 over the strongest non-oracle baseline. Ablations, case studies, and backbone-scaling analyses show that these gains arise from repairing evidence-substrate mismatches through reranking, entity grounding, and structured measurement access, rather than from indiscriminately invoking more reasoning steps, retrieving additional literature, or relying on a particular answer-model scale.

q-bio.QM↗

arXiv2026

From Snapshots to Trajectories: Learning Single-Cell Gene Expression Dynamics via Conditional Flow Matching

Single-cell RNA sequencing (scRNA-seq) provides high-dimensional profiles of cellular states, enabling data-driven modeling of cellular dynamics over time. In practice, time-resolved scRNA-seq is collected at only a few discrete time points as unpaired snapshot populations, leaving substantial temporal gaps. This motivates trajectory inference at unmeasured time points. Existing methods mainly follow two directions, optimal-transport (OT) alignment provides distribution-level matching between observed snapshots, while continuous-time generative models support forecasting via learned dynamics. However, two challenges remain: (i) unpaired snapshots render local transitions between adjacent time points ambiguous, leading to unstable supervision; and (ii) long-horizon prediction relies on repeated integration, where small modeling errors compound and cause distribution drift. To address these challenges, we propose single-cell Flow Matching (scFM), a latent generative framework based on coupling-conditioned flow matching. First, we compute entropically regularized OT couplings between adjacent snapshots and use them to construct soft, weighted flow-matching targets for learning time-dependent velocity fields. Second, we learn bidirectional velocity fields and leverage their consistency to refine couplings and improve temporal coherence under sparse supervision. Third, we introduce distribution-level alignment and latent dynamic regularization to anchor long rollouts and mitigate drift. Experiments on real-world time-series scRNA-seq datasets show that scFM consistently improves distributional prediction performance for both temporal interpolation and extrapolation. Moreover, scFM yields more accurate trajectory reconstruction and temporally coherent visualizations where intermediate time points are absent, indicating a more faithful recovery of underlying temporal gene expression dynamics.

cs.LG↗

arXiv2026

Recent advances in the combination of nonlinearity and exceptional points

The exotic physics emerging at singularities has long attracted intense theoretical and experimental attention. In non-Hermitian systems, exceptional points (EPs), unique spectral singularities, have given rise to a host of intriguing wave phenomena and enabled a broad range of promising applications across diverse physical platforms. Recently, considerable effort has been devoted to combining nonlinearity with exceptional points (EPs) to enable flexible control, overcome the limitations of linear EPs, discover previously unexplored singularities, and reveal novel physical phenomena and application potentials. In this review, we provide a detailed overview of the interplay between nonlinearity and EPs, highlighting key developments such as noise suppression for enhanced sensing, emerging mechanisms for chiral-like state transfer, the realization of optical isolators in nonlinear EP systems, applications including wireless energy transfer and frequency comb generation, among others. We also offer a perspective on future research directions and opportunities in this rapidly evolving field.

physics.optics↗

arXiv2026

Programming active-molecule dynamics via intramolecular nonreciprocity

The dynamics of a self-propelled particle are typically hard-wired by its microscopic construction, limiting the range of behaviors accessible without redesigning the particle itself. Here we show that intramolecular nonreciprocity provides a minimal and versatile mechanism to overcome this constraint. We construct active molecules from short chains of two species of self-propelled particles whose propulsion directions are coupled nonreciprocally according to a prescribed internal sequence. At the single-molecule level, homogeneous sequences exhibit standard persistent random-walk dynamics, whereas heterogeneous sequences produce distinct trajectories inaccessible to either constituent species alone. At the collective level, using motility-induced phase separation (MIPS) as a representative example, we find that modifying the internal sequence shifts the MIPS onset by multiple orders of magnitude in propulsion strength, without altering particle-level interactions. These results demonstrate that intramolecular nonreciprocity among a small set of active components enables sequence-level programmability from single-molecule dynamics to emergent collective behavior, providing a minimal mechanism to encode and control active-matter dynamics across scales.

cond-mat.soft↗

arXiv2026

DeepEra: A Deep Evidence Reranking Agent for Scientific Retrieval-Augmented Generated Question Answering

With the rapid growth of scientific literature, scientific question answering (SciQA) has become increasingly critical for exploring and utilizing scientific knowledge. Retrieval-Augmented Generation (RAG) enhances LLMs by incorporating knowledge from external sources, thereby providing credible evidence for scientific question answering. But existing retrieval and reranking methods remain vulnerable to passages that are semantically similar but logically irrelevant, often reducing factual reliability and amplifying hallucinations.To address this challenge, we propose a Deep Evidence Reranking Agent (DeepEra) that integrates step-by-step reasoning, enabling more precise evaluation of candidate passages beyond surface-level semantics. To support systematic evaluation, we construct SciRAG-SSLI (Scientific RAG - Semantically Similar but Logically Irrelevant), a large-scale dataset comprising about 300K SciQA instances across 10 subjects, constructed from 10M scientific corpus. The dataset combines naturally retrieved contexts with systematically generated distractors to test logical robustness and factual grounding. Comprehensive evaluations confirm that our approach achieves superior retrieval performance compared to leading rerankers. To our knowledge, this work is the first to comprehensively study and empirically validate innegligible SSLI issues in two-stage RAG frameworks.

cs.CL↗

arXiv2026

SciHorizon-GENE: Benchmarking LLM for Life Sciences Inference from Gene Knowledge to Functional Understanding

Large language models (LLMs) have shown growing promise in biomedical research, particularly for knowledge-driven interpretation tasks. However, their ability to reliably reason from gene-level knowledge to functional understanding, a core requirement for knowledge-enhanced cell atlas interpretation, remains largely underexplored. To address this gap, we introduce SciHorizon-GENE, a large-scale gene-centric benchmark constructed from authoritative biological databases. The benchmark integrates curated knowledge for over 190K human genes and comprises more than 540K questions covering diverse gene-to-function reasoning scenarios relevant to cell type annotation, functional interpretation, and mechanism-oriented analysis. Motivated by behavioral patterns observed in preliminary examinations, SciHorizon-GENE evaluates LLMs along four biologically critical perspectives: research attention sensitivity, hallucination tendency, answer completeness, and literature influence, explicitly targeting failure modes that limit the safe adoption of LLMs in biological interpretation pipelines. We systematically evaluate a wide range of state-of-the-art general-purpose and biomedical LLMs, revealing substantial heterogeneity in gene-level reasoning capabilities and persistent challenges in generating faithful, complete, and literature-grounded functional interpretations. Our benchmark establishes a systematic foundation for analyzing LLM behavior at the gene scale and offers insights for model selection and development, with direct relevance to knowledge-enhanced biological interpretation.

q-bio.GN↗

arXiv2025

Observation of Fully Flat Bands in a Photonic Dipolar Kagome Lattice

Flat bands, characterized by zero group velocity and strong energy localization, enable interaction-enhanced phenomena across both quantum and classical systems. Existing photonic flat-band implementations were limited to evanescent-wave systems, specific lattice symmetries, or complex supercell modulations. A simple, universal, and efficient approach to realizing flat bands without dedicated source excitation is to be explored. Here, inspired by geometrically frustrated configurations, we theoretically proposed and experimentally demonstrated threefold-degenerate flat bands by integrating orbital and rotational degrees of freedom in a photonic dipolar kagome lattice. By rotating the dipole orientation, the system exhibits a band flip transition at which point all bands achieve complete flatness and degeneracy across the entire Brillouin zone. In contrast to conventional s-orbital kagome lattices with only a single flat band, our approach flattens the entire band structure, eliminating dispersive modes and enabling compatibility with arbitrary excitations. These results establish a new mechanism for flat-band engineering, offering a tunable strategy for enhancing light-matter interactions and may have applications in compact photonic devices and energy-efficient information processing.

physics.optics↗

arXiv2025

Observation of Arbitrarily Configurable Nonlinear Topological Modes

Nonlinear topology is an emerging field that combines the intrinsic reconfigurability of nonlinear systems with the robustness of topological protection, offering fertile ground for unconventional phenomena and novel applications. Recently, arbitrarily configurable nonlinear topological modes (ANTMs) were proposed, enabling wavefunctions to be configured into arbitrary profiles , and offering greatly enhanced capacity for topological modes and high-throughput topological transport. Here we present the first direct experimental demonstration of ANTMs . These nonlinear topological modes are robust against disorder while also being continuously reshaped and reconfigured in real time through external control. These counterintuitive properties highlight the versatility of arbitrarily morphing nonlinear topological modes and pave the way for highly adaptable topological devices capable of operating reliably across diverse application scenarios, including those involving imperfections, signal variability, and dynamic conditions.

physics.optics↗

arXiv2025

Impact of noise on nonlinear-exceptional-point-based sensors

Nonlinear exceptional points (NEPs), a new type of spectral singularity in nonlinear non-Hermitian systems, are expected to address the noise divergence issue encountered at linear exceptional points and are therefore under the scrutiny of theoretical and experimental investigations. However, concerns have been raised that NEPs may hinder improvements in the signal-to-noise ratio (SNR) of sensors, and there is currently no rigorous theoretical framework to characterize noise effects in NEPs, particularly when accounting for the inherent nonlinear feedback. Here, we develop a new theoretical framework to address the impact of noise on NEP-based sensors, effectively resolving these concerns. The interplay between noise and nonlinearity keeps the average frequency virtually unchanged. In addition, a hidden feedback mechanism limits the increase in detectable uncertainty, together enabling a substantial SNR enhancement at NEPs. Our results resolve the ongoing debate over the SNR of NEPs and lay the groundwork for NEP-based sensor technologies.

physics.optics↗

arXiv2025

SciRerankBench: Benchmarking Rerankers Towards Scientific Retrieval-Augmented Generated LLMs

Scientific literature question answering is a pivotal step towards new scientific discoveries. Recently, \textit{two-stage} retrieval-augmented generated large language models (RAG-LLMs) have shown impressive advancements in this domain. Such a two-stage framework, especially the second stage (reranker), is particularly essential in the scientific domain, where subtle differences in terminology may have a greatly negative impact on the final factual-oriented or knowledge-intensive answers. Despite this significant progress, the potential and limitations of these works remain unexplored. In this work, we present a Scientific Rerank-oriented RAG Benchmark (SciRerankBench), for evaluating rerankers within RAG-LLMs systems, spanning five scientific subjects. To rigorously assess the reranker performance in terms of noise resilience, relevance disambiguation, and factual consistency, we develop three types of question-context-answer (Q-C-A) pairs, i.e., Noisy Contexts (NC), Semantically Similar but Logically Irrelevant Contexts (SSLI), and Counterfactual Contexts (CC). Through systematic evaluation of 13 widely used rerankers on five families of LLMs, we provide detailed insights into their relative strengths and limitations. To the best of our knowledge, SciRerankBench is the first benchmark specifically developed to evaluate rerankers within RAG-LLMs, which provides valuable observations and guidance for their future development.

cs.CL↗

arXiv2025

Soft Graph Clustering for single-cell RNA Sequencing Data

Clustering analysis is fundamental in single-cell RNA sequencing (scRNA-seq) data analysis for elucidating cellular heterogeneity and diversity. Recent graph-based scRNA-seq clustering methods, particularly graph neural networks (GNNs), have significantly improved in tackling the challenges of high-dimension, high-sparsity, and frequent dropout events that lead to ambiguous cell population boundaries. However, their reliance on hard graph constructions derived from thresholded similarity matrices presents challenges:(i) The simplification of intercellular relationships into binary edges (0 or 1) by applying thresholds, which restricts the capture of continuous similarity features among cells and leads to significant information loss.(ii) The presence of significant inter-cluster connections within hard graphs, which can confuse GNN methods that rely heavily on graph structures, potentially causing erroneous message propagation and biased clustering outcomes. To tackle these challenges, we introduce scSGC, a Soft Graph Clustering for single-cell RNA sequencing data, which aims to more accurately characterize continuous similarities among cells through non-binary edge weights, thereby mitigating the limitations of rigid data structures. The scSGC framework comprises three core components: (i) a zero-inflated negative binomial (ZINB)-based feature autoencoder; (ii) a dual-channel cut-informed soft graph embedding module; and (iii) an optimal transport-based clustering optimization module. Extensive experiments across ten datasets demonstrate that scSGC outperforms 13 state-of-the-art clustering models in clustering accuracy, cell type annotation, and computational efficiency. These results highlight its substantial potential to advance scRNA-seq data analysis and deepen our understanding of cellular heterogeneity.

cs.LG↗

arXiv2025

Reinforcement Learning-based Feature Generation Algorithm for Scientific Data

Feature generation (FG) aims to enhance the prediction potential of original data by constructing high-order feature combinations and removing redundant features. It is a key preprocessing step for tabular scientific data to improve downstream machine-learning model performance. Traditional methods face the following two challenges when dealing with the feature generation of scientific data: First, the effective construction of high-order feature combinations in scientific data necessitates profound and extensive domain-specific expertise. Secondly, as the order of feature combinations increases, the search space expands exponentially, imposing prohibitive human labor consumption. Advancements in the Data-Centric Artificial Intelligence (DCAI) paradigm have opened novel avenues for automating feature generation processes. Inspired by that, this paper revisits the conventional feature generation workflow and proposes the Multi-agent Feature Generation (MAFG) framework. Specifically, in the iterative exploration stage, multi-agents will construct mathematical transformation equations collaboratively, synthesize and identify feature combinations ex-hibiting high information content, and leverage a reinforcement learning mechanism to evolve their strategies. Upon completing the exploration phase, MAFG integrates the large language models (LLMs) to interpreta-tively evaluate the generated features of each significant model performance breakthrough. Experimental results and case studies consistently demonstrate that the MAFG framework effectively automates the feature generation process and significantly enhances various downstream scientific data mining tasks.

cs.LG↗

arXiv2025

GCAL: Adapting Graph Models to Evolving Domain Shifts

This paper addresses the challenge of graph domain adaptation on evolving, multiple out-of-distribution (OOD) graphs. Conventional graph domain adaptation methods are confined to single-step adaptation, making them ineffective in handling continuous domain shifts and prone to catastrophic forgetting. This paper introduces the Graph Continual Adaptive Learning (GCAL) method, designed to enhance model sustainability and adaptability across various graph domains. GCAL employs a bilevel optimization strategy. The "adapt" phase uses an information maximization approach to fine-tune the model with new graph domains while re-adapting past memories to mitigate forgetting. Concurrently, the "generate memory" phase, guided by a theoretical lower bound derived from information bottleneck theory, involves a variational memory graph generation module to condense original graphs into memories. Extensive experimental evaluations demonstrate that GCAL substantially outperforms existing methods in terms of adaptability and knowledge retention.

cs.LG↗

arXiv2025

Knowledge-Driven Agentic Scientific Corpus Distillation Framework for Biomedical Large Language Models Training

Corpus distillation for biomedical large language models (LLMs) seeks to address the pressing challenge of insufficient quantity and quality in open-source annotated scientific corpora, which remains a bottleneck for effective LLM training in biomedical research. This paper proposes a knowledge-driven, agentic framework for scientific corpus distillation, tailored explicitly for LLM training in the biomedical domain, addressing the challenge posed by the complex hierarchy of biomedical knowledge. Central to our approach is a collaborative multi-agent architecture, where specialized agents, each guided by the Medical Subject Headings (MeSH) hierarchy, work in concert to autonomously extract, synthesize, and self-evaluate high-quality textual data from vast scientific literature. This agentic framework collectively generates and refines domain-specific question-answer pairs, ensuring comprehensive coverage and consistency with biomedical ontologies while minimizing manual involvement. Extensive experimental results show that language models trained on our multi-agent distilled datasets achieve notable improvements in biomedical question-answering tasks, outperforming both strong life sciences LLM baselines and advanced proprietary models. Notably, our AI-Ready dataset enables Llama3-70B to surpass GPT-4 with MedPrompt and Med-PaLM-2, despite their larger scale. Detailed ablation studies and case analyses further validate the effectiveness and synergy of each agent within the framework, highlighting the potential of multi-agent collaboration in biomedical LLM training.

cs.CL↗

arXiv2025

Collaborative Multi-Agent Reinforcement Learning for Automated Feature Transformation with Graph-Driven Path Optimization

Feature transformation methods aim to find an optimal mathematical feature-feature crossing process that generates high-value features and improves the performance of downstream machine learning tasks. Existing frameworks, though designed to mitigate manual costs, often treat feature transformations as isolated operations, ignoring dynamic dependencies between transformation steps. To address the limitations, we propose TCTO, a collaborative multi-agent reinforcement learning framework that automates feature engineering through graph-driven path optimization. The framework's core innovation lies in an evolving interaction graph that models features as nodes and transformations as edges. Through graph pruning and backtracking, it dynamically eliminates low-impact edges, reduces redundant operations, and enhances exploration stability. This graph also provides full traceability to empower TCTO to reuse high-utility subgraphs from historical transformations. To demonstrate the efficacy and adaptability of our approach, we conduct comprehensive experiments and case studies, which show superior performance across a range of datasets.

cs.LG↗