arXiv subjects

Siqi Wang

Publications and source records attributed to Siqi Wang.

At least 37 records · Page 2Linked to original sources

arXiv2026

PhysFormer: A Physics-Embedded Generative Model for Physically Self-Consistent Spectral Synthesis

In scientific and engineering domains, modeling high-dimensional complex systems governed by partial differential equations (PDEs) remains challenging in terms of physical consistency and numerical stability. However, existing approaches, such as physics-informed neural networks (PINNs), typically rely on known physical fields or coefficients and enforce physical constraints via external loss functions, which can lead to training instability and make it difficult to handle high-dimensional or unobservable scenarios. To this end, we propose PhysFormer, a generative modeling framework that is self-consistent at both the data and physical levels. PhysFormer leverages a low-dimensional, physically interpretable latent space to learn key physical quantities directly from data without requiring known high-dimensional physical field parameters, and embeds the physical process of radiative flux generation within the network to ensure the physical consistency of the generated spectra. In high-dimensional, degenerate inversion tasks, PhysFormer constrains generation within physical limits and enhances spectral fidelity and inversion stability under varying signal-to-noise ratios (SNRs). More broadly, this approach shifts the physical processes from external loss functions into the generative mechanism itself, providing a physically consistent generative modeling paradigm for complex systems involving unknown or unobservable physical quantities.

astro-ph.IM↗

arXiv2026

Noise-Aware Generalization: Robustness to In-Domain Noise and Out-of-Domain Generalization

Methods addressing Learning with Noisy Labels (LNL) and multi-source Domain Generalization (DG) use training techniques to improve downstream task performance in the presence of label noise or domain shifts, respectively. Prior work often explores these tasks in isolation, and the limited work that does investigate their intersection, which we refer to as Noise-Aware Generalization (NAG), only benchmarks existing methods without also proposing an approach to reduce its effect. We find that this is likely due, in part, to the new challenges that arise when exploring NAG, which does not appear in LNL or DG alone. For example, we show that the effectiveness of DG methods is compromised in the presence of label noise, making them largely ineffective. Similarly, LNL methods often overfit to easy-to-learn domains as they confuse domain shifts for label noise. Instead, we propose Domain Labels for Noise Detection (DL4ND), the first direct method developed for NAG which uses our observation that noisy samples that may appear indistinguishable within a single domain often show greater variation when compared across domains. We find DL4ND outperforms DG and LNL methods, including their combinations, even when simplifying the NAG challenge by using domain labels to isolate domain shifts from noise. Performance gains up to 12.5% over seven diverse datasets with three noise types demonstrates DL4ND's ability to generalize to a wide variety of settings.

cs.LG↗

arXiv2026

Generative Diffusion Augmentation with Quantum-Enhanced Discrimination for Medical Image Diagnosis

In biomedical engineering, artificial intelligence has become a pivotal tool for enhancing medical diagnostics, particularly in medical image classification tasks such as detecting pneumonia from chest X-rays and breast cancer screening. However, real-world medical datasets frequently exhibit severe class imbalance, where positive samples substantially outnumber negative samples, leading to biased models with low recall rates for minority classes. This imbalance not only compromises diagnostic accuracy but also poses clinical misdiagnosis risks. To address this challenge, we propose SDA-QEC (Simplified Diffusion Augmentation with Quantum-Enhanced Classification), an innovative framework that integrates simplified diffusion-based data augmentation with quantum-enhanced feature discrimination. Our approach employs a lightweight diffusion augmentor to generate high-quality synthetic samples for minority classes, rebalancing the training distribution. Subsequently, a quantum feature layer embedded within MobileNetV2 architecture enhances the model's discriminative capability through high-dimensional feature mapping in Hilbert space. Comprehensive experiments on coronary angiography image classification demonstrate that SDA-QEC achieves 98.33% accuracy, 98.78% AUC, and 98.33% F1-score, significantly outperforming classical baselines including ResNet18, MobileNetV2, DenseNet121, and VGG16. Notably, our framework simultaneously attains 98.33% sensitivity and 98.33% specificity, achieving a balanced performance critical for clinical deployment. The proposed method validates the feasibility of integrating generative augmentation with quantum-enhanced modeling in real-world medical imaging tasks, offering a novel research pathway for developing highly reliable medical AI systems in small-sample, highly imbalanced, and high-risk diagnostic scenarios.

cs.CV↗

arXiv2026

Domain Generalization with Quantum Enhancement for Medical Image Classification: A Lightweight Approach for Cross-Center Deployment

Medical image artificial intelligence models often achieve strong performance in single-center or single-device settings, yet their effectiveness frequently deteriorates in real-world cross-center deployment due to domain shift, limiting clinical generalizability. To address this challenge, we propose a lightweight domain generalization framework with quantum-enhanced collaborative learning, enabling robust generalization to unseen target domains without relying on real multi-center labeled data. Specifically, a MobileNetV2-based domain-invariant encoder is constructed and optimized through three key components: (1) multi-domain imaging shift simulation using brightness, contrast, sharpening, and noise perturbations to emulate heterogeneous acquisition conditions; (2) domain-adversarial training with gradient reversal to suppress domain-discriminative features; and (3) a lightweight quantum feature enhancement layer that applies parameterized quantum circuits for nonlinear feature mapping and entanglement modeling. In addition, a test-time adaptation strategy is employed during inference to further alleviate distribution shifts. Experiments on simulated multi-center medical imaging datasets demonstrate that the proposed method significantly outperforms baseline models without domain generalization or quantum enhancement on unseen domains, achieving reduced domain-specific performance variance and improved AUC and sensitivity. These results highlight the clinical potential of quantum-enhanced domain generalization under constrained computational resources and provide a feasible paradigm for hybrid quantum--classical medical imaging systems.

cs.CV↗

arXiv2026

A Lightweight Medical Image Classification Framework via Self-Supervised Contrastive Learning and Quantum-Enhanced Feature Modeling

Intelligent medical image analysis is essential for clinical decision support but is often limited by scarce annotations, constrained computational resources, and suboptimal model generalization. To address these challenges, we propose a lightweight medical image classification framework that integrates self-supervised contrastive learning with quantum-enhanced feature modeling. MobileNetV2 is employed as a compact backbone and pretrained using a SimCLR-style self-supervised paradigm on unlabeled images. A lightweight parameterized quantum circuit (PQC) is embedded as a quantum feature enhancement module, forming a hybrid classical-quantum architecture, which is subsequently fine-tuned on limited labeled data. Experimental results demonstrate that, with only approximately 2-3 million parameters and low computational cost, the proposed method consistently outperforms classical baselines without self-supervised learning or quantum enhancement in terms of Accuracy, AUC, and F1-score. Feature visualization further indicates improved discriminability and representation stability. Overall, this work provides a practical and forward-looking solution for high-performance medical artificial intelligence under resource-constrained settings.

cs.CV↗

arXiv2026

A Lightweight Brain-Inspired Machine Learning Framework for Coronary Angiography: Hybrid Neural Representation and Robust Learning Strategies

Background: Coronary angiography (CAG) is a cornerstone imaging modality for assessing coronary artery disease and guiding interventional treatment decisions. However, in real-world clinical settings, angiographic images are often characterized by complex lesion morphology, severe class imbalance, label uncertainty, and limited computational resources, posing substantial challenges to conventional deep learning approaches in terms of robustness and generalization.Methods: The proposed framework is built upon a pretrained convolutional neural network to construct a lightweight hybrid neural representation. A selective neural plasticity training strategy is introduced to enable efficient parameter adaptation. Furthermore, a brain-inspired attention-modulated loss function, combining Focal Loss with label smoothing, is employed to enhance sensitivity to hard samples and uncertain annotations. Class-imbalance-aware sampling and cosine annealing with warm restarts are adopted to mimic rhythmic regulation and attention allocation mechanisms observed in biological neural systems.Results: Experimental results demonstrate that the proposed lightweight brain-inspired model achieves strong and stable performance in binary coronary angiography classification, yielding competitive accuracy, recall, F1-score, and AUC metrics while maintaining high computational efficiency.Conclusion: This study validates the effectiveness of brain-inspired learning mechanisms in lightweight medical image analysis and provides a biologically plausible and deployable solution for intelligent clinical decision support under limited computational resources.

cs.CV↗

arXiv2026

Beyond Feature Mapping GAP: Integrating Real HDRTV Priors for Superior SDRTV-to-HDRTV Conversion

The rise of HDR-WCG display devices has highlighted the need to convert SDRTV to HDRTV, as most video sources are still in SDR. Existing methods primarily focus on designing neural networks to learn a single-style mapping from SDRTV to HDRTV. However, the limited information in SDRTV and the diversity of styles in real-world conversions render this process an ill-posed problem, thereby constraining the performance and generalization of these methods. Inspired by generative approaches, we propose a novel method for SDRTV to HDRTV conversion guided by real HDRTV priors. Despite the limited information in SDRTV, introducing real HDRTV as reference priors significantly constrains the solution space of the originally high-dimensional ill-posed problem. This shift transforms the task from solving an unreferenced prediction problem to making a referenced selection, thereby markedly enhancing the accuracy and reliability of the conversion process. Specifically, our approach comprises two stages: the first stage employs a Vector Quantized Generative Adversarial Network to capture HDRTV priors, while the second stage matches these priors to the input SDRTV content to recover realistic HDRTV outputs. We evaluate our method on public datasets, demonstrating its effectiveness with significant improvements in both objective and subjective metrics across real and synthetic datasets.

eess.IV↗

arXiv2026

ASCNet: Research on all-sky camera images classification at the Muztagh-ata site

Cloud coverage is one of the crucial elements of site testing in astronomy. All-sky camera (ASC) images are beneficial for our research on cloud coverage. In this paper, we propose ASCNet, an innovative model specifically designed for classifying nighttime ASC images collected at the Muztagh-ata site from 2022 March to 2024 June. ASCNet integrates ResNet34 with an ASCModule, which employs Depthwise Dilated Convolution and embeds lightweight Squeeze-and-Excitation attention within its branches to extract fine-grained texture information from the luminance channel. The data set is partitioned by category, with 70% of images assigned to the training set and 30% to the test set. The model's performance is assessed by comparing its predictions on the test set with manually annotated labels, yielding a consistency rate of 92.7%. All evaluation metrics of ASCNet are as follows: Accuracy 92.66%, Precision 83.26%, Recall 84.25%, and F1-Score 83.67%, and both ablation and comparative experiments demonstrate significant superiority over other models. A confusion matrix is utilized to analyze the differences between manual classification and model classification. The statistical results demonstrate the model's excellent classification performance and its robust generalization ability, illustrating that ASCNet has potential for application in future astronomical image classifications.

astro-ph.IM↗

arXiv2025

Superconductivity suppression and bilayer decoupling in Pr substituted YBa$_2$Cu$_3$O$_{7-δ}$

The mechanism behind superconductivity suppression induced by Pr substitutions in YBa$_2$Cu$_3$O$_{7-δ}$ (YBCO) has been a mystery since its discovery: in spite of being isovalent to Y$^{3+}$ with a small magnetic moment, it is the only rare-earth element that has a dramatic impact on YBCO's superconducting properties. Using angle-resolved photoemission spectroscopy (ARPES) and DFT+$U$ calculations, we uncover how Pr substitution modifies the low-energy electronic structure of YBCO. Contrary to the prevailing Fehrenbacher-Rice (FR) and Liechtenstein-Mazin (LM) models, the low energy electronic structure contains no signature of any $f$-electron hybridization or new states. Yet, strong electron doping is observed primarily on the antibonding Fermi surface. Meanwhile, we reveal major electronic structure modifications to Cu-derived states with increasing Pr substitution: a pronounced CuO$_2$ bilayer decoupling and an enhanced CuO chain hopping, implying indirect electron-release pathways beyond simple 4$f$ state ionization. Our results challenge the long-standing FR/LM mechanism and establish Pr substituted YBCO as a potential platform for exploring correlation-driven phenomena in coupled 1D-2D systems.

cond-mat.supr-con↗

arXiv2025

LaF-GRPO: In-Situ Navigation Instruction Generation for the Visually Impaired via GRPO with LLM-as-Follower Reward

Navigation instruction generation for visually impaired (VI) individuals (NIG-VI) is critical yet relatively underexplored. This study focuses on generating precise, in-situ, step-by-step navigation instructions that are practically usable for VI users. Specifically, we propose LaF-GRPO (LLM-as-Follower GRPO), where an LLM simulates VI user responses to navigation instructions, thereby providing feedback rewards to guide the post-training of a Vision-Language Model (VLM). This enhances instruction accuracy and usability while reducing costly real-world data collection needs. To address the scarcity of dedicated benchmarks in this field, we introduce NIG4VI, a 27k-sample open-source dataset to facilitate training and evaluation. It comprises diverse navigation scenarios with accurate spatial coordinates, supporting detailed and open-ended in-situ instruction generation. Experiments on NIG4VI demonstrate the effectiveness of LaF-GRPO through quantitative metrics (e.g., Zero-(LaF-GRPO) boosts BLEU 14\%; SFT+(LaF-GRPO) METEOR 0.542 vs. GPT-4o 0.323), and qualitative analysis further confirms that our method yields more intuitive and safer instructions.

cs.CL↗

arXiv2025

MEML-GRPO: Heterogeneous Multi-Expert Mutual Learning for RLVR Advancement

Recent advances demonstrate that reinforcement learning with verifiable rewards (RLVR) significantly enhances the reasoning capabilities of large language models (LLMs). However, standard RLVR faces challenges with reward sparsity, where zero rewards from consistently incorrect candidate answers provide no learning signal, particularly in challenging tasks. To address this, we propose Multi-Expert Mutual Learning GRPO (MEML-GRPO), an innovative framework that utilizes diverse expert prompts as system prompts to generate a broader range of responses, substantially increasing the likelihood of identifying correct solutions. Additionally, we introduce an inter-expert mutual learning mechanism that facilitates knowledge sharing and transfer among experts, further boosting the model's performance through RLVR. Extensive experiments across multiple reasoning benchmarks show that MEML-GRPO delivers significant improvements, achieving an average performance gain of 4.89% with Qwen and 11.33% with Llama, effectively overcoming the core limitations of traditional RLVR methods.

cs.AI↗

arXiv2025

CitySeeker: How Do VLMS Explore Embodied Urban Navigation With Implicit Human Needs?

Vision-Language Models (VLMs) have made significant progress in explicit instruction-based navigation; however, their ability to interpret implicit human needs (e.g., "I am thirsty") in dynamic urban environments remains underexplored. This paper introduces CitySeeker, a novel benchmark designed to assess VLMs' spatial reasoning and decision-making capabilities for exploring embodied urban navigation to address implicit needs. CitySeeker includes 6,440 trajectories across 8 cities, capturing diverse visual characteristics and implicit needs in 7 goal-driven scenarios. Extensive experiments reveal that even top-performing models (e.g., Qwen2.5-VL-32B-Instruct) achieve only 21.1% task completion. We find key bottlenecks in error accumulation in long-horizon reasoning, inadequate spatial cognition, and deficient experiential recall. To further analyze them, we investigate a series of exploratory strategies-Backtracking Mechanisms, Enriching Spatial Cognition, and Memory-Based Retrieval (BCR), inspired by human cognitive mapping's emphasis on iterative observation-reasoning cycles and adaptive path optimization. Our analysis provides actionable insights for developing VLMs with robust spatial intelligence required for tackling "last-mile" navigation challenges.

cs.AI↗

arXiv2025

RLHFSpec: Breaking the Efficiency Bottleneck in RLHF Training via Adaptive Drafting

Reinforcement Learning from Human Feedback (RLHF) is an important fine-tuning technique for large language models (LLMs) and comprises three stages: generation, inference, and training. The generation stage generates samples that are then used to infer learnable experiences for training. We observe that the generation stage is the bottleneck of the entire execution process and consider it a key point for optimization. Specifically, we realize the first attempt to integrate speculative decoding into the RLHF generation stage and propose RLHFSpec, an RLHF system that accelerates generation execution with efficient speculative decoding and sample reallocation. To fully exploit the performance potential provided by speculative decoding, especially dealing with the dynamic workload of the generation stage, RLHFSpec proposes a workload-aware drafting strategy selection mechanism, which selects the near-optimal strategy by jointly considering the verification cost and the number of accepted tokens. Moreover, RLHFSpec also proposes sample reallocation to fully utilize the GPU resources, and optimizes it with an efficient sample migration mechanism. The experimental results show that the RLHFSpec can achieve higher throughput in the generation stage compared to state-of-the-art works. Moreover, due to the effective alleviation of the generation bottleneck, RLHFSpec also shows significant performance speedup in the entire RLHF execution.

cs.LG↗

arXiv2025

Scaling and Transferability of Annealing Strategies in Large Language Model Training

Learning rate scheduling is crucial for training large language models, yet understanding the optimal annealing strategies across different model configurations remains challenging. In this work, we investigate the transferability of annealing dynamics in large language model training and refine a generalized predictive framework for optimizing annealing strategies under the Warmup-Steady-Decay (WSD) scheduler. Our improved framework incorporates training steps, maximum learning rate, and annealing behavior, enabling more efficient optimization of learning rate schedules. Our work provides a practical guidance for selecting optimal annealing strategies without exhaustive hyperparameter searches, demonstrating that smaller models can serve as reliable proxies for optimizing the training dynamics of larger models. We validate our findings on extensive experiments using both Dense and Mixture-of-Experts (MoE) models, demonstrating that optimal annealing ratios follow consistent patterns and can be transferred across different training configurations.

cs.LG↗

arXiv2025

AniME: Adaptive Multi-Agent Planning for Long Animation Generation

We present AniME, a director-oriented multi-agent system for automated long-form anime production, covering the full workflow from a story to the final video. The director agent keeps a global memory for the whole workflow, and coordinates several downstream specialized agents. By integrating customized Model Context Protocol (MCP) with downstream model instruction, the specialized agent adaptively selects control conditions for diverse sub-tasks. AniME produces cinematic animation with consistent characters and synchronized audio visual elements, offering a scalable solution for AI-driven anime creation.

cs.AI↗

arXiv2025

Post-Completion Learning for Language Models

Current language model training paradigms typically terminate learning upon reaching the end-of-sequence ( ) token, overlooking the potential learning opportunities in the post-completion space. We propose Post-Completion Learning (PCL), a novel training framework that systematically utilizes the sequence space after model output completion, to enhance both the reasoning and self-evaluation abilities. PCL enables models to continue generating self-assessments and reward predictions during training, while maintaining efficient inference by stopping at the completion point. To fully utilize this post-completion space, we design a white-box reinforcement learning method: let the model evaluate the output content according to the reward rules, then calculate and align the score with the reward functions for supervision. We implement dual-track SFT to optimize both reasoning and evaluation capabilities, and mixed it with RL training to achieve multi-objective hybrid optimization. Experimental results on different datasets and models demonstrate consistent improvements over traditional SFT and RL methods. Our method provides a new technical path for language model training that enhances output quality while preserving deployment efficiency.

cs.CL↗

arXiv2025

Dichotomy of flat bands in the van der Waals ferromagnet Fe$_5$GeTe$_2$

Quantum materials with bands of narrow bandwidth near the Fermi level represent a promising platform for exploring a diverse range of fascinating physical phenomena, as the high density of states within the small energy window often enables the emergence of many-body physics. On one hand, flat bands can arise from strong Coulomb interactions that localize atomic orbitals. On the other hand, quantum destructive interference can quench the electronic kinetic energy. Although both have a narrow bandwidth, the two types of flat bands should exhibit very distinct spectral properties arising from their distinctive origins. So far, the two types of flat bands have only been realized in very different material settings and chemical environments, preventing a direct comparison. Here, we report the observation of the two types of flat bands within the same material system--an above-room-temperature van der Waals ferromagnet, Fe$_{5-x}$GeTe$_2$, distinguishable by a switchable iron site order. The contrasting nature of the flat bands is also identified by the remarkably distinctive temperature-evolution of the spectral features, indicating that one arises from electron correlations in the Fe(1) site-disordered phase, while the other geometrical frustration in the Fe(1) site-ordered phase. Our results therefore provide a direct juxtaposition of the distinct formation mechanism of flat bands in quantum materials, and an avenue for understanding the distinctive roles flat bands play in the presence of magnetism, topology, and lattice geometrical frustration, utilizing sublattice ordering as a key control parameter.

cond-mat.str-el↗

arXiv2025

Refactoring Deep Learning Code: A Study of Practices and Unsatisfied Tool Needs

With the rapid development of deep learning, the implementation of intricate algorithms and substantial data processing have become standard elements of deep learning projects. As a result, the code has become progressively complex as the software evolves, which is difficult to maintain and understand. Existing studies have investigated the impact of refactoring on software quality within traditional software. However, the insight of code refactoring in the context of deep learning is still unclear. This study endeavors to fill this knowledge gap by empirically examining the current state of code refactoring in deep learning realm, and practitioners' views on refactoring. We first manually analyzed the commit history of five popular and well-maintained deep learning projects (e.g., PyTorch). We mined 4,921 refactoring practices in historical commits and measured how different types and elements of refactoring operations are distributed and found that refactoring operation types' distribution in deep learning projects is different from it in traditional Java software. We then surveyed 159 practitioners about their views of code refactoring in deep learning projects and their expectations of current refactoring tools. The result of the survey showed that refactoring research and the development of related tools in the field of deep learning are crucial for improving project maintainability and code quality, and that current refactoring tools do not adequately meet the needs of practitioners. Lastly, we provided our perspective on the future advancement of refactoring tools and offered suggestions for developers' development practices.

cs.SE↗