arXiv subjects

Wei Xiong

Publications and source records attributed to Wei Xiong.

At least 55 records · Page 3Linked to original sources

arXiv2025

Gene regulatory network inference algorithm based on spectral signed directed graph convolution

Accurately reconstructing Gene Regulatory Networks (GRNs) is crucial for understanding gene functions and disease mechanisms. Single-cell RNA sequencing (scRNA-seq) technology provides vast data for computational GRN reconstruction. Since GRNs are ideally modeled as signed directed graphs to capture activation/inhibition relationships, the most intuitive and reasonable approach is to design feature extractors based on the topological structure of GRNs to extract structural features, then combine them with biological characteristics for research. However, traditional spectral graph convolution struggles with this representation. Thus, we propose MSGRNLink, a novel framework that explicitly models GRNs as signed directed graphs and employs magnetic signed Laplacian convolution. Experiments across simulated and real datasets demonstrate that MSGRNLink outperforms all baseline models in AUROC. Parameter sensitivity analysis and ablation studies confirmed its robustness and the importance of each module. In a bladder cancer case study, MSGRNLink predicted more known edges and edge signs than benchmark models, further validating its biological relevance.

q-bio.MN↗

arXiv2025

Reinforce-Ada: An Adaptive Sampling Framework under Non-linear RL Objectives

Reinforcement learning (RL) for large language model reasoning is frequently hindered by signal loss, a phenomenon where standard uniform sampling with small group sizes fails to uncover informative learning signals for difficult prompts. We demonstrate that this collapse is a statistical artifact of undersampling rather than an inherent model limitation. To address this systematically, we introduce a theoretical framework based on optimizing a non-linear RL objective (e.g., log-likelihood). We show that this objective naturally induces a weighted gradient estimator that prioritizes difficult prompts, which can be robustly realized through adaptive sampling. Guided by this framework, we propose Reinforce-Ada, a family of algorithms that dynamically allocates inference budgets based on prompt difficulty, effectively scaling up RL compute to where it is needed most. Unlike passive filtering methods that discard low-signal prompts, Reinforce-Ada actively invests compute to recover them. We introduce two efficient realizations: an estimation-based approach and a model-free sequential sampling approach. Extensive experiments across multiple benchmarks show that Reinforce-Ada significantly outperforms uniform baselines like GRPO, recovering lost signals and accelerating convergence by up to $2\times$ while maintaining the same total inference budget. Code is available at https://github.com/RLHFlow/Reinforce-Ada.

cs.LG↗

arXiv2025

polyRETRO: a Language Model Approach to predict Polymerization Class and Monomer(s) for a Target Polymer

While machine learning has transformed polymer design by enabling rapid property prediction and candidate generation, translating these designs into experimentally realizable materials remains a critical challenge. Traditionally, the synthesis of target polymers has relied heavily on expert intuition and prior experience. The lack of automated retrosynthetic tools to assist chemists, limit the rapid practical impact of data-driven polymer discovery. To expedite lab-scale validation and beyond, we present a retrosynthetic framework that leverages large language models (LLMs) to guide polymer synthesis. Our approach, which we call polyRETRO, involves two key steps: 1) predicting the most likely polymerization reaction class of a target polymer and 2) identifying the underlying chemical transformation templates and the corresponding monomers, using primarily natural-language based constructs. This LLM-driven framework enables direct retrosynthetic analysis given just the target polymer SMILES string. polyRETRO constitutes a initial step towards a scalable, interpretable, and generalizable approach to bridge the gap between computational design and experimental synthesis.

cond-mat.soft↗

arXiv2025

Controllable Bistability in Dual-Fiber Optical Trap in Air

The dual-fiber optical trap, owing to its high sensitivity and facile miniaturization, holds significant actual application value in fields such as high-precision metrology of mechanical quantities and biological manipulation. The positional stability of the trapped particle is pivotal to system performance, directly setting the measurement noise floor and operational precision. In this work, we observe bistability and hysteresis in the axial equilibrium position of a 10-um diameter SiO2 microsphere. This bistability arises from optical interference between the fiber ends and the microsphere, creating multiple potential wells. Experimental results demonstrate that the microsphere's transition rate can be effectively modulated through precise control of the trapping laser power. Furthermore, the incorporation of transverse misalignment has effectively eradicated bistability, thereby substantially improving positional stability throughout the entire optical trapping region. This suppression successfully reduced the system's residual positional uncertainty to the thermal noise limit. Consequently, this research will enhance the precision of microparticle manipulation and the sensitivity of sensing in dual-fiber optical trap systems.

physics.optics↗

arXiv2025

Quasinormal modes of spontaneous scalarized Kerr black holes

Recent studies have shown that rotating black holes can undergo spontaneous scalarization, leading to deviations from general relativity in the strong-field regime. We present the first nonperturbative calculation of the quasinormal modes (QNMs) of scalarized Kerr black holes in Einstein-scalar-Gauss-Bonnet gravity, without assuming small spin or weak coupling. Our results reveal a universal splitting of the fundamental $l=m=2$ mode into axial-led, polar-led, and scalar-led branches, breaking the isospectrality characteristic of Kerr black holes. This splitting offers distinct signatures in the ringdown phase of gravitational wave signals and provides a new avenue to test gravity in the strong-field regime. Our findings open the possibility of probing beyond-GR physics using precision measurements of black hole ringdowns in upcoming gravitational wave observations.

gr-qc↗

arXiv2025

AI-assisted design of chemically recyclable polymers for food packaging

Polymer packaging plays a crucial role in food preservation but poses major challenges in recycling and environmental persistence. To address the need for sustainable, high-performance alternatives, we employed a polymer informatics workflow to identify single- and multi-layer drop-in replacements for polymer-based packaging materials. Machine learning (ML) models, trained on carefully curated polymer datasets, predicted eight key properties across a library of approximately 7.4 million ring-opening polymerization (ROP) polymers generated by virtual forward synthesis (VFS). Candidates were prioritized by the enthalpy of polymerization, a critical metric for chemical recyclability. This screening yielded thousands of promising candidates, demonstrating the feasibility of replacing diverse packaging architectures. We then experimentally validated poly(p-dioxanone) (poly-PDO), an existing ROP polymer whose barrier performance had not been previously reported. Validation showed that poly-PDO exhibits strong water barrier performance, mechanical and thermal properties consistent with predictions, and excellent chemical recyclability (95% monomer recovery), thereby meeting the design targets and underscoring its potential for sustainable packaging. These findings highlight the power of informatics-driven approaches to accelerate the discovery of sustainable polymers by uncovering opportunities in both existing and novel chemistries.

cond-mat.soft↗

arXiv2025

AI-Driven Design of poly(ethylene terephthalate)-replacement copolymers

Poly(ethylene terephthalate) (PET), a widely used thermoplastic in packaging, textiles, and engineering applications, is valued for its strength, clarity, and chemical resistance. Increasing environmental impact concerns and regulatory pressures drive the search for alternatives with comparable or superior performance. We present an AI-driven polymer design pipeline employing virtual forward synthesis (VFS) to generate PET-replacement copolymers. Inspired by the esterification route of PET synthesis, we systematically combined a down-selected set of Toxic Substances Control Act (TSCA)-listed monomers to create 12,100 PET-like polymers. Machine learning models predicted glass transition temperature (Tg), band gap, and tendency to crystallize, for all designs. Multi-objective screening identified 1,108 candidates predicted to match or exceed PET in $T_{\rm g}$ and band gap, including the ``rediscovery'' of other known commercial PET-alternate polymers (e.g., PETG, Tritan, Ecozen) that provide retrospective validation of our design pipeline, demonstrating a capability to rapidly design experimentally feasible polymers at a scale. Furthermore, selected, entirely new (previously unknown) candidates designed here have been synthesized and characterized, providing a definitive validation of the design framework.

cond-mat.soft↗

arXiv2025

Hybrid Cavity-Magnon Optomechanics: Tailoring Bipartite and Tripartite Macroscopic Entanglement

Cavity optomechanics, providing an inherently nonlinear interaction between photons and phonons, have shown enomerous potential in generating macroscopic quantum entanglement. Here we propose to realize diverse bipartite and tripartite entanglement in cavity-magnon optomechanics. By introducing magnons to standard cavity optomechanics, not only tunable optomechanical entanglement and magnon-magnon entanglement can be achieved, but also flexible tripartite entanglement including magnon-photon-phonon entanglement, magnon-magnon-photon and -phonon entanglement can be generated. Moreover, optimal bipartite and tripartite entanglement can be achieved by tuning parameters. We further show that all entanglement can be enhanced via engineering the magnon-photon coupling, and is proven to be robust against the bath temperature within the survival temperature. Besides, we find that the optomechanical entanglement can be protected or restored by bad magnons with large decay rate, while other entanglement is severely reduced. The results indicate that our proposal provides a novel avenue to explore and control tunable macroscopic quantum effects in hybrid cavity-magnon optomechanics.

quant-ph↗

arXiv2025

An Encoder-Decoder Foundation Chemical Language Model for Generative Polymer Design

Traditional machine learning has advanced polymer discovery, yet direct generation of chemically valid and synthesizable polymers without exhaustive enumeration remains a challenge. Here we present polyT5, an encoder-decoder chemical language model based on the T5 architecture, trained to understand and generate polymer structures. polyT5 enables both property prediction and the targeted generation of polymers conditioned on desired property values. We demonstrate its utility for dielectric polymer design, seeking candidates with dielectric constant >3, bandgap >4 eV, and glass transition temperature >400 K, alongside melt-processability and solubility requirements. From over 20,000 generated promising candidates, one was experimentally synthesized and validated, showing strong agreement with predictions. To further enhance usability, we integrated polyT5 within an agentic AI framework that couples it with a general-purpose LLM, allowing natural language interaction for property prediction and generative design. Together, these advances establish a versatile and accessible framework for accelerated polymer discovery.

cond-mat.mtrl-sci↗

arXiv2025

Determination of all complete mappings of F_{q^2} of the form aX^{3q}+bX^{2q+1}+cX^{q+2}+dX^3

For each prime power q, we determine all polynomials over F_{q^2} of the form f(X) := aX^{3q}+bX^{2q+1}+cX^{q+2}+dX^3 which induce complete mappings of F_{q^2}, in the sense that each of the functions x --> f(x) and x --> f(x)+x permutes F_{q^2}. This is the first result in the literature which classifies the complete mappings among some class of polynomials with arbitrarily large degree over finite fields of arbitrary characteristic. We also determine all permutation polynomials over F_{q^2} of the form X^{q+2}+bX^q+cX, and all permutations of (F_q)^2 induced by maps of the form (x,y) --> (x^3-exy^2-ax-by, y^3-cx-dy) where either e=0 or 3|q. The latter results add to the small number of results in the literature classifying all permutations induced by maps of prescribed forms.

math.NT↗

arXiv2025

polyBART: A Chemical Linguist for Polymer Property Prediction and Generative Design

Designing polymers for targeted applications and accurately predicting their properties is a key challenge in materials science owing to the vast and complex polymer chemical space. While molecular language models have proven effective in solving analogous problems for molecular discovery, similar advancements for polymers are limited. To address this gap, we propose polyBART, a language model-driven polymer discovery capability that enables rapid and accurate exploration of the polymer design space. Central to our approach is Pseudo-polymer SELFIES (PSELFIES), a novel representation that allows for the transfer of molecular language models to the polymer space. polyBART is, to the best of our knowledge, the first language model capable of bidirectional translation between polymer structures and properties, achieving state-of-the-art results in property prediction and design of novel polymers for electrostatic energy storage. Further, polyBART is validated through a combination of both computational and laboratory experiments. We report what we believe is the first successful synthesis and validation of a polymer designed by a language model, predicted to exhibit high thermal degradation temperature and confirmed by our laboratory measurements. Our work presents a generalizable strategy for adapting molecular language models to the polymer space and introduces a polymer foundation model, advancing generative polymer design that may be adapted for a variety of applications.

cond-mat.soft↗

arXiv2025

RAMS: Residual-based adversarial-gradient moving sample method for scientific machine learning in solving partial differential equations

Physics-informed neural networks (PINNs) and neural operators, two leading scientific machine learning (SciML) paradigms, have emerged as powerful tools for solving partial differential equations (PDEs). Although increasing the training sample size generally enhances network performance, it also increases computational costs for physics-informed or data-driven training. To address this trade-off, different sampling strategies have been developed to sample more points in regions with high PDE residuals. However, existing sampling methods are computationally demanding for high-dimensional problems, such as high-dimensional PDEs or operator learning tasks. Here, we propose a residual-based adversarial-gradient moving sample (RAMS) method, which moves samples according to the adversarial gradient direction to maximize the PDE residual via gradient-based optimization. RAMS can be easily integrated into existing sampling methods. Extensive experiments, ranging from PINN applied to high-dimensional PDEs to physics-informed and data-driven operator learning problems, have been conducted to demonstrate the effectiveness of RAMS. Notably, RAMS represents the first efficient adaptive sampling approach for operator learning, marking a significant advancement in the SciML field.

cs.CE↗

arXiv2025

Semiparametric model averaging for high-dimensional quantile regression with nonignorable nonresponse

Model averaging has demonstrated superior performance for ensemble forecasting in high-dimensional framework, its extension to incomplete datasets remains a critical but underexplored challenge. Moreover, identifying the parsimonious model through averaging procedure in quantile regression demands urgent methodological innovation. In this paper, we propose a novel model averaging method for high-dimensional quantile regression with nonignorable missingness. The idea is to relax the parametric constraint on the conditional distribution of respondents, which is constructed through the two-phase scheme: (i) a semiparametric likelihood-based estimation for the missing mechanism, and (ii) a semiparametric weighting procedure to combine candidate models. One of pivotal advantages is our SMA estimator can asymptotically concentrate on the optimally correct model when the candidate set involves at least one correct model. Theoretical results show that the estimator achieves asymptotic optimality even under complex missingness conditions. Empirical conclusions illustrate the efficiency of the method.

stat.ME↗

arXiv2025

StepWiser: Stepwise Generative Judges for Wiser Reasoning

As models increasingly leverage multi-step reasoning strategies to solve complex problems, supervising the logical validity of these intermediate steps has become a critical research challenge. Process reward models address this by providing step-by-step feedback, but current approaches have two major drawbacks: they typically function as classifiers without providing explanations, and their reliance on supervised fine-tuning with static datasets limits generalization. Inspired by recent advances, we reframe stepwise reward modeling from a classification task to a reasoning task itself. We thus propose a generative judge that reasons about the policy model's reasoning steps (i.e., meta-reasons), outputting thinking tokens before delivering a final verdict. Our model, StepWiser, is trained by reinforcement learning using relative outcomes of rollouts. We show it provides (i) better judgment accuracy on intermediate steps than existing methods; (ii) can be used to improve the policy model at training time; and (iii) improves inference-time search.

cs.AI↗

arXiv2025

DIVE: Taming DINO for Subject-Driven Video Editing

Building on the success of diffusion models in image generation and editing, video editing has recently gained substantial attention. However, maintaining temporal consistency and motion alignment still remains challenging. To address these issues, this paper proposes DINO-guided Video Editing (DIVE), a framework designed to facilitate subject-driven editing in source videos conditioned on either target text prompts or reference images with specific identities. The core of DIVE lies in leveraging the powerful semantic features extracted from a pretrained DINOv2 model as implicit correspondences to guide the editing process. Specifically, to ensure temporal motion consistency, DIVE employs DINO features to align with the motion trajectory of the source video. For precise subject editing, DIVE incorporates the DINO features of reference images into a pretrained text-to-image model to learn Low-Rank Adaptations (LoRAs), effectively registering the target subject's identity. Extensive experiments on diverse real-world videos demonstrate that our framework can achieve high-quality editing results with robust motion consistency, highlighting the potential of DINO to contribute to video editing. Project page: https://dino-video-editing.github.io

cs.CV↗

arXiv2025

LLM Alignment as Retriever Optimization: An Information Retrieval Perspective

Large Language Models (LLMs) have revolutionized artificial intelligence with capabilities in reasoning, coding, and communication, driving innovation across industries. Their true potential depends on effective alignment to ensure correct, trustworthy and ethical behavior, addressing challenges like misinformation, hallucinations, bias and misuse. While existing Reinforcement Learning (RL)-based alignment methods are notoriously complex, direct optimization approaches offer a simpler alternative. In this work, we introduce a novel direct optimization approach for LLM alignment by drawing on established Information Retrieval (IR) principles. We present a systematic framework that bridges LLM alignment and IR methodologies, mapping LLM generation and reward models to IR's retriever-reranker paradigm. Building on this foundation, we propose LLM Alignment as Retriever Preference Optimization (LarPO), a new alignment method that enhances overall alignment quality. Extensive experiments validate LarPO's effectiveness with 38.9 % and 13.7 % averaged improvement on AlpacaEval2 and MixEval-Hard respectively. Our work opens new avenues for advancing LLM alignment by integrating IR foundations, offering a promising direction for future research.

cs.CL↗

arXiv2025

Strong and noise-tolerant entanglement in dissipative optomechanics

Macroscopic entanglement, as a critical quantum resource in quantum information science, has been extensively studied in coherent optomechanics over the past decades. However, entanglement in dissipative optomechanics, where the cavity linewidth depends on the position of the mechanical resonator, remains largely unexplored. In this work, we investigate quantum entanglement in a dissipative optomechanical system realized by a Michelson-Sagnac interferometer with a movable membrane. This configuration enables the switching between coherent and dissipative optomechanical couplings at will. With experimentally feasible parameters, we demonstrate that the steady-state mechanical displacement exhibits a nonlinear (linear) dependence on the driving power under coherent (dissipative) coupling. Furthermore, we show that the quantum entanglement generated via dissipative coupling is significantly stronger and more robust to noise than that generated via coherent coupling. When both coherent and dissipative couplings are simultaneously present, the entanglement is weakened due to quantum interference. Our results indicate that dissipative optomechanical coupling can be a promising route for engineering strong and noise-resilient quantum entanglement.

quant-ph↗

arXiv2025

When do molecular polaritons behave like optical filters?

This review outlines several linear optical effects featured by molecular polaritons arising in the collective strong light-matter coupling regime. Under weak laser irradiation and when the single-molecule light-matter coupling can be neglected (often in the limit when the number of molecules per photon mode is large), we show that the excited-state molecular dynamics under collective strong coupling can be exactly replicated without the cavity using a shaped (or ``filtered'') laser, whose field amplitude is enhanced by the cavity quality factor, shining on the bare molecules. As a consequence, the absorption within a cavity can be understood as the overlap between the polariton transmission and the bare molecular absorption, suggesting that polaritons act in part as optical filters. This framework demystifies and provides a straightforward explanation for a large class of experiments and theoretical models in molecular polaritonics, highlighting that the same effects can be achieved without the cavity with shaped laser pulses. With a few modifications, this simple conceptual picture can also be adapted to understand the incoherent nonlinear response of polaritonic systems. This review establishes a clear distinction between polaritonic phenomena that can be fully explained through classical linear optics and those that require a quantum electrodynamics approach. It also highlights the need to differentiate between effects that necessitate polaritons (i.e., hybrid light-matter states) and those that can occur in the weak coupling regime. We further discuss that certain quantum optical effects like fluorescence can be partially described as optical filtering, whereas some others like cavity-induced Raman scattering go beyond this. Further exploration in these areas is needed to uncover novel polaritonic phenomena beyond optical filtering.

physics.chem-ph↗