Searcharxiv⌕ Search

arXiv subjects

Yifan Chen

Publications and source records attributed to Yifan Chen.

At least 109 records · Page 6Linked to original sources
arXiv2025

Stable Derivative Free Gaussian Mixture Variational Inference for Bayesian Inverse Problems

This paper is concerned with the approximation of probability distributions known up to normalization constants, with a focus on Bayesian inference for large-scale inverse problems in scientific computing. In this context, key challenges include costly repeated evaluations of forward models, multimodality, and inaccessible gradients for the forward model. To address them, we develop a variational inference framework that combines Fisher-Rao natural gradient with specialized quadrature rules to enable derivative free updates of Gaussian mixture variational families. The resulting method, termed Derivative Free Gaussian Mixture Variational Inference (DF-GMVI), guarantees covariance positivity and affine invariance, offering a stable and efficient framework for approximating complex posterior distributions. The effectiveness of DF-GMVI is demonstrated through numerical experiments on challenging scenarios, including distributions with multiple modes, infinitely many modes, and curved modes in spaces with up to 100 dimensions. The method's practicality is further demonstrated in a large-scale application, where it successfully recovers the initial conditions of the Navier-Stokes equations from solution data at positive times.

cs.LG↗
arXiv2025

Ground Calibration Result of the Wide-field X-ray Telescope (WXT) onboard the Einstein Probe

We report on results of the on-ground X-ray calibration of the Wide-field X-ray Telescope (WXT) built from novel lobster-eye micro-pore optics, onboard the Einstein Probe (EP) satellite. To fully characterize the instrumental performance and properties, a series of tests and calibrations have been carried out at different levels of devices, assemblies and the complete module before the launch of EP. In this paper, we present the calibration results of three flight model modules (FM1, FM5 and FM11) obtained during their end-to-end module calibration experiments carried out at the 100-m X-ray Test Facility (100XF) of IHEP, CAS. Measurements of the Point Spread Function (PSF), effective area, and energy response were performed for multiple incident directions and several characteristic X-ray emission line energies. Specifically, the distributions of the PSF and effective areas are found to be roughly uniform across the FoV, in large agreement with the prediction of lobster-eye optics. Their energy dependence behavior aligns well with theoretical predictions and Monte Carlo simulations. At 1.25 keV, the full width at half maximum (FWHM) of the focal spot is in range of 3-7 arcmin (a median of 4.2) and the effective area in range of 2-3 $cm^2$. Noticeably, the flight model instruments demonstrate a $\sim1.5$ arcmin spatial resolution improvement over the previously launched Lobster Eye Imager for Astronomy. The properties of the complementary metal-oxide semiconductor (CMOS) sensors were also calibrated. The gain coefficients are in range of 6.4-6.9 eV/DN. The energy resolutions are in range of 120-140 eV at 1.25 keV, meeting design requirements. These calibration results have been ingested into the first version of calibration database (CALDB) and applied to the analysis of the scientific data acquired by WXT after the launch of EP.

astro-ph.IM↗
arXiv2025

Partition-wise Graph Filtering: A Unified Perspective Through the Lens of Graph Coarsening

Filtering-based graph neural networks (GNNs) constitute a distinct class of GNNs that employ graph filters to handle graph-structured data, achieving notable success in various graph-related tasks. Conventional methods adopt a graph-wise filtering paradigm, imposing a uniform filter across all nodes, yet recent findings suggest that this rigid paradigm struggles with heterophilic graphs. To overcome this, recent works have introduced node-wise filtering, which assigns distinct filters to individual nodes, offering enhanced adaptability. However, a fundamental gap remains: a comprehensive framework unifying these two strategies is still absent, limiting theoretical insights into the filtering paradigms. Moreover, through the lens of Contextual Stochastic Block Model, we reveal that a synthesis of graph-wise and node-wise filtering provides a sufficient solution for classification on graphs exhibiting both homophily and heterophily, suggesting the risk of excessive parameterization and potential overfitting with node-wise filtering. To address the limitations, this paper introduces Coarsening-guided Partition-wise Filtering (CPF). CPF innovates by performing filtering on node partitions. The method begins with structure-aware partition-wise filtering, which filters node partitions obtained via graph coarsening algorithms, and then performs feature-aware partition-wise filtering, refining node embeddings via filtering on clusters produced by $k$-means clustering over features. In-depth analysis is conducted for each phase of CPF, showing its superiority over other paradigms. Finally, benchmark node classification experiments, along with a real-world graph anomaly detection application, validate CPF's efficacy and practical utility.

cs.LG↗
arXiv2025

Synergistic interplay of morphology and metabolic activity rule response to CAR-T cells in B-cell lymphomas

Cellular immunotherapies are one of the mainstream cancer treatments unveiling the power of the patient's immune system to fight tumors. CAR T-cell therapy, based on genetically engineered T cells, has demonstrated significant potential in treating hematological malignancies, including B-cell lymphomas. This treatment has complex longitudinal dynamics due to the interplay of different T-cell phenotypes (e.g. effector and memory), the expansion of the drug and the cytotoxic effect on both normal and cancerous B-cells, the exhaustion of the immune cells, the tumor immunosupressive environments, and more. Thus, the outcome of the therapy is not yet well understood leading to a variety of responses ranging from sustained complete responses, different types of partial responses, or no response at all. We developed a mechanistic model for the interaction between CAR T- and cancerous B-cells, accounting for the role of the tumor morphology and metabolic status. The simulations showed that lesions with irregular shapes and high proliferation could contribute to long term progression by potentially increasing their immunosuppressive capabilities impairing CAR T-cell efficacy. We analyzed 18F-FDG PET/CT imaging data from 63 relapsed/refractory diffuse large B-cell lymphoma receiving CAR T-cells, quantifying radiomic features including tumor sphericity and lesion aggressiveness through standardized uptake values (SUV). Statistical analyses revealed significant correlations between these metrics and progression-free survival (PFS), emphasizing that individual lesions with complex morphology and elevated metabolism play a critical role in shaping long-term treatment outcomes. We demonstrated the potential of using data-driven mathematical models in finding molecular-imaging based biomarkers to identify lymphoma patients treated with CAR T-cell therapy having higher risk of disease progression.

q-bio.PE↗
arXiv2025

Dynamical Lensing Tomography of Black Hole Ringdown

Strong gravitational lensing occurs when photons pass through the vicinity of a black hole. We investigate this phenomenon in the context of a gravitational-wave event, specifically when a black hole is settling into its final state. The deflection angle of photons mimics the ringdown pattern of the gravitational wave at intermediate times. At late times it has an inverse cubic dependence on observation time. The deviation angle increases exponentially as photons approach the photon ring orbit, reflecting its unstable nature. Our findings are directly applicable to imaging scenarios involving stars against the background of compact binaries, or circumbinary accretion disks, particularly during the merger of two black holes.

gr-qc↗
arXiv2025

Simultaneous Resonant and Broadband Detection of Ultralight Dark Matter and High-Frequency Gravitational Waves via Cavities and Circuits

Electromagnetic resonant systems, such as cavities and LC circuits, are widely used to detect ultralight boson dark matter and high-frequency gravitational waves. However, the narrow bandwidth of single-mode resonators necessitates multiple scan steps to cover broad frequency ranges. By incorporating a network of auxiliary modes via beam-splitter-type and non-degenerate parametric couplings, we enable broadband detection with an effective bandwidth of each scan matching the order of the resonant frequency, while maintaining a strong signal response. In heterodyne upconversion detection, where a background cavity mode transitions into another due to a potential background source, multiple orders of the source frequency can be probed with high sensitivity without tuning the cavity frequency. Consequently, our method allows for significantly deeper exploration of the parameter space within the same integration time compared to single-mode detection.

hep-ph↗
arXiv2025

Embedding Empirical Distributions for Computing Optimal Transport Maps

Distributional data have become increasingly prominent in modern signal processing, highlighting the necessity of computing optimal transport (OT) maps across multiple probability distributions. Nevertheless, recent studies on neural OT methods predominantly focused on the efficient computation of a single map between two distributions. To address this challenge, we introduce a novel approach to learning transport maps for new empirical distributions. Specifically, we employ the transformer architecture to produce embeddings from distributional data of varying length; these embeddings are then fed into a hypernetwork to generate neural OT maps. Various numerical experiments were conducted to validate the embeddings and the generated OT maps. The model implementation and the code are provided on https://github.com/jiangmingchen/HOTET.

cs.LG↗
arXiv2025

Enhancing Diagnostic Accuracy in Rare and Common Fundus Diseases with a Knowledge-Rich Vision-Language Model

Previous foundation models for fundus images were pre-trained with limited disease categories and knowledge base. Here we introduce a knowledge-rich vision-language model (RetiZero) that leverages knowledge from more than 400 fundus diseases. For RetiZero's pretraining, we compiled 341,896 fundus images paired with texts, sourced from public datasets, ophthalmic literature, and online resources, encompassing a diverse range of diseases across multiple ethnicities and countries. RetiZero exhibits remarkable performance in several downstream tasks, including zero-shot disease recognition, image-to-image retrieval, AI-assisted clinical diagnosis,few-shot fine-tuning, and internal- and cross-domain disease identification. In zero-shot scenarios, RetiZero achieves Top-5 accuracies of 0.843 for 15 diseases and 0.756 for 52 diseases. For image retrieval, it achieves Top-5 scores of 0.950 and 0.886 for the same sets, respectively. AI-assisted clinical diagnosis results show that RetiZero's Top-3 zero-shot performance surpasses the average of 19 ophthalmologists from Singapore, China, and the United States. RetiZero substantially enhances clinicians' accuracy in diagnosing fundus diseases, in particularly rare ones. These findings underscore the value of integrating the RetiZero into clinical settings, where various fundus diseases are encountered.

eess.IV↗
arXiv2025

On Stringy E-functions and the Non-negativity Conjecture for Determinantal Varieties

We compute the stringy E-functions of determinantal varieties and establish that the stringy E-function of a determinantal variety coincides with the E-function of the product of a Grassmannian and an affine space. Furthermore, a similar result holds for the projectivization of determinantal varieties. As an application, we verify the non-negativity conjecture for the stringy Hodge numbers of these varieties.

math.AG↗
arXiv2025

Forward Ray Tracing and Hot Spots in Kerr Spacetime

Hotspots, often characterized as pointlike emissions, frequently appear near black holes with significantly enhanced luminosity compared to the surrounding accretion flow. Notably, such hotspots are regularly observed near the black hole at the center of the Milky Way. Light rays emitted from these sources follow complex trajectories around the black hole before reaching distinct locations on the observer's image plane. Precisely resolving both direct emissions and their higher-order images--despite the latter's intensity suppression--is essential for extracting detailed spacetime information, including the black hole's mass, spin, and inclination angle. To improve the accuracy and efficiency of hotspot modeling, we develop a forward ray tracing method based on the analytic integral solution of Kerr geodesics, leveraging conserved quantities. Our approach traces geodesics from a given emission point near the black hole to a distant observer, effectively capturing multiple images with a tailored parametrization scheme for root-finding. By perturbing these geodesics, we map finite-size emissions to distinct regions on the image plane, enabling the quantification of image shapes and amplification rates. This method not only enhances the identification of strongly lensed photons from black holes but also enables efficient spacetime tomography and hotspot localization, leveraging observations from the Event Horizon Telescope and its upcoming next-generation upgrades.

gr-qc↗
arXiv2025

SuperGPQA: Scaling LLM Evaluation across 285 Graduate Disciplines

Large language models (LLMs) have demonstrated remarkable proficiency in mainstream academic disciplines such as mathematics, physics, and computer science. However, human knowledge encompasses over 200 specialized disciplines, far exceeding the scope of existing benchmarks. The capabilities of LLMs in many of these specialized fields-particularly in light industry, agriculture, and service-oriented disciplines-remain inadequately evaluated. To address this gap, we present SuperGPQA, a comprehensive benchmark that evaluates graduate-level knowledge and reasoning capabilities across 285 disciplines. Our benchmark employs a novel Human-LLM collaborative filtering mechanism to eliminate trivial or ambiguous questions through iterative refinement based on both LLM responses and expert feedback. Our experimental results reveal significant room for improvement in the performance of current state-of-the-art LLMs across diverse knowledge domains (e.g., the reasoning-focused model DeepSeek-R1 achieved the highest accuracy of 61.82% on SuperGPQA), highlighting the considerable gap between current model capabilities and artificial general intelligence. Additionally, we present comprehensive insights from our management of a large-scale annotation process, involving over 80 expert annotators and an interactive Human-LLM collaborative system, offering valuable methodological guidance for future research initiatives of comparable scope.

cs.CL↗
arXiv2025

Large model enhanced computational ghost imaging

Ghost imaging (GI) achieves 2D image reconstruction through high-order correlation of 1D bucket signals and 2D light field information, particularly demonstrating enhanced detection sensitivity and high-quality image reconstruction via efficient photon collection in scattering media. Recent investigations have established that deep learning (DL) can substantially enhance the ghost imaging reconstruction quality. Furthermore, with the emergence of large models like SDXL, GPT-4, etc., the constraints of conventional DL in parameters and architecture have been transcended, enabling models to comprehensively explore relationships among all distinct positions within feature sequences. This paradigm shift has significantly advanced the capability of DL in restoring severely degraded and low-resolution imagery, making it particularly advantageous for noise-robust image reconstruction in GI applications. In this paper, we propose the first large imaging model with 1.4 billion parameters that incorporates the physical principles of GI (GILM). The proposed GILM implements a skip connection mechanism to mitigate gradient explosion challenges inherent in deep architectures, ensuring sufficient parametric capacity to capture intricate correlations among object single-pixel measurements. Moreover, GILM leverages multi-head attention mechanism to learn spatial dependencies across pixel points during image reconstruction, facilitating the extraction of comprehensive object information for subsequent reconstruction. We validated the effectiveness of GILM through a series of experiments, including simulated object imaging, imaging objects in free space, and imaging object located 52 meters away in underwater environment. The experimental results show that GILM effectively analyzes the fluctuation trends of the collected signals, thereby optimizing the recovery of the object's image from the acquired data.

eess.IV↗
arXiv2025

ResMoE: Space-efficient Compression of Mixture of Experts LLMs via Residual Restoration

Mixture-of-Experts (MoE) Transformer, the backbone architecture of multiple phenomenal language models, leverages sparsity by activating only a fraction of model parameters for each input token. The sparse structure, while allowing constant time costs, results in space inefficiency: we still need to load all the model parameters during inference. We introduce ResMoE, an innovative MoE approximation framework that utilizes Wasserstein barycenter to extract a common expert (barycenter expert) and approximate the residuals between this barycenter expert and the original ones. ResMoE enhances the space efficiency for inference of large-scale MoE Transformers in a one-shot and data-agnostic manner without retraining while maintaining minimal accuracy loss, thereby paving the way for broader accessibility to large language models. We demonstrate the effectiveness of ResMoE through extensive experiments on Switch Transformer, Mixtral, and DeepSeekMoE models. The results show that ResMoE can reduce the number of parameters in an expert by up to 75% while maintaining comparable performance. The code is available at https://github.com/iDEA-iSAIL-Lab-UIUC/ResMoE.

cs.LG↗
arXiv2025

Unraveling and Mitigating Retriever Inconsistencies in Retrieval-Augmented Large Language Models

Although Retrieval-Augmented Large Language Models (RALMs) demonstrate their superiority in terms of factuality, they do not consistently outperform the original retrieval-free Language Models (LMs). Our experiments reveal that this example-level performance inconsistency exists not only between retrieval-augmented and retrieval-free LM but also among different retrievers. To understand this phenomenon, we investigate the degeneration behavior of RALMs and theoretically decompose it into four categories. Further analysis based on our decomposition reveals that the innate difference in knowledge sources and the unpredictable degeneration of the reader model contribute most to the inconsistency. Drawing from our analysis, we introduce Ensemble of Retrievers (EoR), a trainable framework that can adaptively retrieve from different knowledge sources and effectively decrease unpredictable reader errors. Our experiments on Open Domain Question Answering show that EoR substantially improves performance over the RALM with a single retriever by considerably reducing inconsistent behaviors.

cs.AI↗
arXiv2025

LexRAG: Benchmarking Retrieval-Augmented Generation in Multi-Turn Legal Consultation Conversation

Retrieval-augmented generation (RAG) has proven highly effective in improving large language models (LLMs) across various domains. However, there is no benchmark specifically designed to assess the effectiveness of RAG in the legal domain, which restricts progress in this area. To fill this gap, we propose LexRAG, the first benchmark to evaluate RAG systems for multi-turn legal consultations. LexRAG consists of 1,013 multi-turn dialogue samples and 17,228 candidate legal articles. Each sample is annotated by legal experts and consists of five rounds of progressive questioning. LexRAG includes two key tasks: (1) Conversational knowledge retrieval, requiring accurate retrieval of relevant legal articles based on multi-turn context. (2) Response generation, focusing on producing legally sound answers. To ensure reliable reproducibility, we develop LexiT, a legal RAG toolkit that provides a comprehensive implementation of RAG system components tailored for the legal domain. Additionally, we introduce an LLM-as-a-judge evaluation pipeline to enable detailed and effective assessment. Through experimental analysis of various LLMs and retrieval methods, we reveal the key limitations of existing RAG systems in handling legal consultation conversations. LexRAG establishes a new benchmark for the practical application of RAG systems in the legal domain, with its code and data available at https://github.com/CSHaitao/LexRAG.

cs.CL↗
arXiv2025

CaseGen: A Benchmark for Multi-Stage Legal Case Documents Generation

Legal case documents play a critical role in judicial proceedings. As the number of cases continues to rise, the reliance on manual drafting of legal case documents is facing increasing pressure and challenges. The development of large language models (LLMs) offers a promising solution for automating document generation. However, existing benchmarks fail to fully capture the complexities involved in drafting legal case documents in real-world scenarios. To address this gap, we introduce CaseGen, the benchmark for multi-stage legal case documents generation in the Chinese legal domain. CaseGen is based on 500 real case samples annotated by legal experts and covers seven essential case sections. It supports four key tasks: drafting defense statements, writing trial facts, composing legal reasoning, and generating judgment results. To the best of our knowledge, CaseGen is the first benchmark designed to evaluate LLMs in the context of legal case document generation. To ensure an accurate and comprehensive evaluation, we design the LLM-as-a-judge evaluation framework and validate its effectiveness through human annotations. We evaluate several widely used general-domain LLMs and legal-specific LLMs, highlighting their limitations in case document generation and pinpointing areas for potential improvement. This work marks a step toward a more effective framework for automating legal case documents drafting, paving the way for the reliable application of AI in the legal field. The dataset and code are publicly available at https://github.com/CSHaitao/CaseGen.

cs.CL↗
arXiv2025

Black Holes as Fermion Factories

Ultralight bosons near rotating black holes can undergo significant growth through superradiant energy extraction, potentially reaching field values close to the Planck scale and transforming black holes into effective transducers for these fields. The interaction between boson fields and fermions may lead to parametric production or Schwinger pair production of fermions, with efficiencies significantly exceeding those of perturbative decay processes. Additionally, the spatial gradients of scalar clouds and the electric components of vector clouds can accelerate fermions, resulting in observable fluxes. This study considers both Standard Model neutrinos and dark sector fermions, which could contribute to boosted dark matter. Energy loss due to fermion emissions can potentially quench the exponential growth of the cloud, leading to a saturated state. This dynamic provides a framework for establishing limits on boson-neutrino interactions, previously constrained by neutrino self-interaction considerations. In the saturation phase, boson clouds have the capacity to accelerate fermions to TeV energies, producing fluxes that surpass those from atmospheric neutrinos near black holes. These fluxes open new avenues for observations through high-energy neutrino detectors like IceCube, as well as through dark matter direct detection efforts focused on targeted black holes.

hep-ph↗