arXiv subjects

Yanchun Liang

Publications and source records attributed to Yanchun Liang.

At least 19 recordsLinked to original sources

arXiv2026

ReliableRAG: Combating Misinformation in Retrieval-Augmented Generation via Reliability-Guided Reasoning Chains

Retrieval-Augmented Generation (RAG) has emerged as a powerful architecture for Question Answering (QA) by integrating external information into Large Language Models (LLMs). However, false, inaccurate, and misleading information in news and social media poses a serious challenge to real-world RAG systems, especially in multi-hop QA, where complex multi-step reasoning can be misled by even a single deceptive misinformation segment in the retrieved documents. Existing approaches mainly rely on implicit alignment or explicit regulation, but their limited ability to assess fine-grained information reliability makes them vulnerable to deceptive misinformation that is semantically relevant to the question yet factually incorrect, leading to erroneous answers. To address this limitation, we propose ReliableRAG, which, to the best of our knowledge, is the first reliability-driven framework that mitigates deceptive misinformation in multi-hop QA through fine-grained evaluation of individual triples. ReliableRAG first extracts information segments from source documents and represents them as structured triples. It then quantifies triple reliability by combining query-triple semantic relevance with triple credibility, retaining only the top-$K$ reliable and non-redundant triples. Based on these refined triples, ReliableRAG autoregressively constructs robust reasoning chains to consolidate trustworthy evidence and filter deceptive misinformation, producing accurate answers faithful to reliable information. Experiments on three multi-hop QA datasets show that ReliableRAG outperforms existing methods, substantially improving the factual reliability and robustness of RAG systems under deceptive misinformation injection.

cs.CL↗

arXiv2026

LLM-PDESR: Robust PDE Discovery via Subdomain Weighted Residuals and LLM-Guided Symbolic Hypothesis Generation

Discovering governing partial differential equations (PDEs) from noisy observational data is a fundamental challenge in scientific machine learning. Traditional symbolic regression (SR) methods often struggle to identify accurate equations within vast combinatorial search spaces, largely due to their inability to incorporate essential domain-specific prior knowledge. Furthermore, reliance on pointwise evaluations and discrete finite differences inherently amplifies high-frequency noise, creating deceptive fitness landscapes that derail the optimization process. To resolve these bottlenecks, we propose LLM-PDESR, a framework that integrates the structural hypothesis generation of Large Language Models (LLMs) with a mathematically rigorous evaluation environment. By employing C^4-continuous quintic splines for robust differentiation and subdomain weighted residuals as natural low-pass filters, our approach effectively mitigates the fitness landscape distortion that plagues existing methods. A Pareto-driven feedback loop then enables the LLM to iteratively refine candidate equations, balancing predictive accuracy with structural parsimony. We evaluate LLM-PDESR on 23 canonical PDEs and five structurally novel equations (including a multivariate system) specifically designed to preclude dataset memorization and test true discovery capabilities. Demonstrating real-world applicability, the framework successfully extracts a consistent structural skeleton for an interpretable 1D dynamical surrogate (1D-CACE) directly from noisy ERA5 reanalysis data. Extensive experiments and out-of-distribution testing confirm that LLM-PDESR significantly outperforms state-of-the-art methodologies in structural recovery, noise resilience, and the avoidance of spurious complexity and equation bloat.

cs.LG↗

arXiv2026

RHVI-FDD: A Hierarchical Decoupling Framework for Low-Light Image Enhancement

Low-light images often suffer from severe noise, detail loss, and color distortion, which hinder downstream multimedia analysis and retrieval tasks. The degradation in low-light images is complex: luminance and chrominance are coupled, while within the chrominance, noise and details are deeply entangled, preventing existing methods from simultaneously correcting color distortion, suppressing noise, and preserving fine details. To tackle the above challenges, we propose a novel hierarchical decoupling framework (RHVI-FDD). At the macro level, we introduce the RHVI transform, which mitigates the estimation bias caused by input noise and enables robust luminance-chrominance decoupling. At the micro level, we design a Frequency-Domain Decoupling (FDD) module with three branches for further feature separation. Using the Discrete Cosine Transform, we decompose chrominance features into low, mid, and high-frequency bands that predominantly represent global tone, local details, and noise components, which are then processed by tailored expert networks in a divide-and-conquer manner and fused via an adaptive gating module for content-aware fusion. Extensive experiments on multiple low-light datasets demonstrate that our method consistently outperforms existing state-of-the-art approaches in both objective metrics and subjective visual quality.

cs.CV↗

arXiv2025

Towards Efficient Few-shot Graph Neural Architecture Search via Partitioning Gradient Contribution

To address the weight coupling problem, certain studies introduced few-shot Neural Architecture Search (NAS) methods, which partition the supernet into multiple sub-supernets. However, these methods often suffer from computational inefficiency and tend to provide suboptimal partitioning schemes. To address this problem more effectively, we analyze the weight coupling problem from a novel perspective, which primarily stems from distinct modules in succeeding layers imposing conflicting gradient directions on the preceding layer modules. Based on this perspective, we propose the Gradient Contribution (GC) method that efficiently computes the cosine similarity of gradient directions among modules by decomposing the Vector-Jacobian Product during supernet backpropagation. Subsequently, the modules with conflicting gradient directions are allocated to distinct sub-supernets while similar ones are grouped together. To assess the advantages of GC and address the limitations of existing Graph Neural Architecture Search methods, which are limited to searching a single type of Graph Neural Networks (Message Passing Neural Networks (MPNNs) or Graph Transformers (GTs)), we propose the Unified Graph Neural Architecture Search (UGAS) framework, which explores optimal combinations of MPNNs and GTs. The experimental results demonstrate that GC achieves state-of-the-art (SOTA) performance in supernet partitioning quality and time efficiency. In addition, the architectures searched by UGAS+GC outperform both the manually designed GNNs and those obtained by existing NAS methods. Finally, ablation studies further demonstrate the effectiveness of all proposed methods.

cs.LG↗

arXiv2024

Chemical abundances of 20 barium stars from the OHP spectra

Based on the high resolution and high signal-to-noise spectra, we derived the chemical abundances of 20 elements for 20 barium (Ba-) stars. For the first time, the detailed abundances of four sample stars, namely HD 92482, HD 150430, HD 151101 and HD 177304 have been analyzed. Additionally, Ba element abundance has been measured using high resolution spectra for the first time in six of the other 16 sample stars. Based on the [s/Fe] ratios, the Ba-unknown star HD 115927 can be classified as a strong Ba-star, while the Ba-likely star HD 160538 can be categorized into a mild Ba-star. Consequently, our sample comprises three strong and 17 mild Ba-stars. The light odd-Z metal elements and Fe-peak elements exhibit near-solar abundances. The [α/Fe] ratios demonstrate decreasing trends with increasing metallicity. Moreover, the abundances of n-capture elements show significant enhancements in different degrees. Using a threshold of the signed distances to the solar r-process abundance pattern ds = 0.6, we find that all of our sample stars are normal Ba-stars, indicating that the enhancements of s-process elements should be attributed to material transfer from their companions. We compare the observed n-capture patterns of sample stars with the FRUITY models, and estimate the mass of the Thermally-Pulsing Asymptotic Giant Branch stars that previously contaminated the Ba-stars. The models with low masses can successfully explain the observations. From a kinematic point of view, we note that most of our sample stars are linked with the thin disk, while HD 130255 may be associated with the thick disk.

astro-ph.SR↗

arXiv2024

Reinforcement Learning-based Non-Autoregressive Solver for Traveling Salesman Problems

The Traveling Salesman Problem (TSP) is a well-known combinatorial optimization problem with broad real-world applications. Recently, neural networks have gained popularity in this research area because as shown in the literature, they provide strong heuristic solutions to TSPs. Compared to autoregressive neural approaches, non-autoregressive (NAR) networks exploit the inference parallelism to elevate inference speed but suffer from comparatively low solution quality. In this paper, we propose a novel NAR model named NAR4TSP, which incorporates a specially designed architecture and an enhanced reinforcement learning strategy. To the best of our knowledge, NAR4TSP is the first TSP solver that successfully combines RL and NAR networks. The key lies in the incorporation of NAR network output decoding into the training process. NAR4TSP efficiently represents TSP encoded information as rewards and seamlessly integrates it into reinforcement learning strategies, while maintaining consistent TSP sequence constraints during both training and testing phases. Experimental results on both synthetic and real-world TSPs demonstrate that NAR4TSP outperforms five state-of-the-art models in terms of solution quality, inference speed, and generalization to unseen scenarios.

cs.AI↗

arXiv2024

Meta-GPS++: Enhancing Graph Meta-Learning with Contrastive Learning and Self-Training

Node classification is an essential problem in graph learning. However, many models typically obtain unsatisfactory performance when applied to few-shot scenarios. Some studies have attempted to combine meta-learning with graph neural networks to solve few-shot node classification on graphs. Despite their promising performance, some limitations remain. First, they employ the node encoding mechanism of homophilic graphs to learn node embeddings, even in heterophilic graphs. Second, existing models based on meta-learning ignore the interference of randomness in the learning process. Third, they are trained using only limited labeled nodes within the specific task, without explicitly utilizing numerous unlabeled nodes. Finally, they treat almost all sampled tasks equally without customizing them for their uniqueness. To address these issues, we propose a novel framework for few-shot node classification called Meta-GPS++. Specifically, we first adopt an efficient method to learn discriminative node representations on homophilic and heterophilic graphs. Then, we leverage a prototype-based approach to initialize parameters and contrastive learning for regularizing the distribution of node embeddings. Moreover, we apply self-training to extract valuable information from unlabeled nodes. Additionally, we adopt S$^2$ (scaling & shifting) transformation to learn transferable knowledge from diverse tasks. The results on real-world datasets show the superiority of Meta-GPS++. Our code is available here.

cs.LG↗

arXiv2024

Troublemaker Learning for Low-Light Image Enhancement

Low-light image enhancement (LLIE) restores the color and brightness of underexposed images. Supervised methods suffer from high costs in collecting low/normal-light image pairs. Unsupervised methods invest substantial effort in crafting complex loss functions. We address these two challenges through the proposed TroubleMaker Learning (TML) strategy, which employs normal-light images as inputs for training. TML is simple: we first dim the input and then increase its brightness. TML is based on two core components. First, the troublemaker model (TM) constructs pseudo low-light images from normal images to relieve the cost of pairwise data. Second, the predicting model (PM) enhances the brightness of pseudo low-light images. Additionally, we incorporate an enhancing model (EM) to further improve the visual performance of PM outputs. Moreover, in LLIE tasks, characterizing global element correlations is important because more information on the same object can be captured. CNN cannot achieve this well, and self-attention has high time complexity. Accordingly, we propose Global Dynamic Convolution (GDC) with O(n) time complexity, which essentially imitates the partial calculation process of self-attention to formulate elementwise correlations. Based on the GDC module, we build the UGDC model. Extensive quantitative and qualitative experiments demonstrate that UGDC trained with TML can achieve competitive performance against state-of-the-art approaches on public datasets. The code is available at https://github.com/Rainbowman0/TML_LLIE.

eess.IV↗

arXiv2023

Incorporating Surprisingly Popular Algorithm and Euclidean Distance-based Adaptive Topology into PSO

While many Particle Swarm Optimization (PSO) algorithms only use fitness to assess the performance of particles, in this work, we adopt Surprisingly Popular Algorithm (SPA) as a complementary metric in addition to fitness. Consequently, particles that are not widely known also have the opportunity to be selected as the learning exemplars. In addition, we propose a Euclidean distance-based adaptive topology to cooperate with SPA, where each particle only connects to k number of particles with the shortest Euclidean distance during each iteration. We also introduce the adaptive topology into heterogeneous populations to better solve large-scale problems. Specifically, the exploration sub-population better preserves the diversity of the population while the exploitation sub-population achieves fast convergence. Therefore, large-scale problems can be solved in a collaborative manner to elevate the overall performance. To evaluate the performance of our method, we conduct extensive experiments on various optimization problems, including three benchmark suites and two real-world optimization problems. The results demonstrate that our Euclidean distance-based adaptive topology outperforms the other widely adopted topologies and further suggest that our method performs significantly better than state-of-the-art PSO variants on small, medium, and large-scale problems.

cs.NE↗

arXiv2021

Crossed-Time Delay Neural Network for Speaker Recognition

Time Delay Neural Network (TDNN) is a well-performing structure for DNN-based speaker recognition systems. In this paper we introduce a novel structure Crossed-Time Delay Neural Network (CTDNN) to enhance the performance of current TDNN. Inspired by the multi-filters setting of convolution layer from convolution neural network, we set multiple time delay units each with different context size at the bottom layer and construct a multilayer parallel network. The proposed CTDNN gives significant improvements over original TDNN on both speaker verification and identification tasks. It outperforms in VoxCeleb1 dataset in verification experiment with a 2.6% absolute Equal Error Rate improvement. In few shots condition CTDNN reaches 90.4% identification accuracy, which doubles the identification accuracy of original TDNN. We also compare the proposed CTDNN with another new variant of TDNN, FTDNN, which shows that our model has a 36% absolute identification accuracy improvement under few shots condition and can better handle training of a larger batch in a shorter training time, which better utilize the calculation resources. The code of the new model is released at https://github.com/chenllliang/CTDNN

eess.AS↗

arXiv2021

Neural Architecture Search based on Cartesian Genetic Programming Coding Method

Neural architecture search (NAS) is a hot topic in the field of automated machine learning and outperforms humans in designing neural architectures on quite a few machine learning tasks. Motivated by the natural representation form of neural networks by the Cartesian genetic programming (CGP), we propose an evolutionary approach of NAS based on CGP, called CGPNAS, to solve sentence classification task. To evolve the architectures under the framework of CGP, the operations such as convolution are identified as the types of function nodes of CGP, and the evolutionary operations are designed based on Evolutionary Strategy. The experimental results show that the searched architectures are comparable with the performance of human-designed architectures. We verify the ability of domain transfer of our evolved architectures. The transfer experimental results show that the accuracy deterioration is lower than 2-5%. Finally, the ablation study identifies the Attention function as the single key function node and the linear transformations along could keep the accuracy similar with the full evolved architectures, which is worthy of investigation in the future.

cs.NE↗

arXiv2020

Deep Learning Analysis and Age Prediction from Shoeprints

Human walking and gaits involve several complex body parts and are influenced by personality, mood, social and cultural traits, and aging. These factors are reflected in shoeprints, which in turn can be used to predict age, a problem not systematically addressed using any computational approach. We collected 100,000 shoeprints of subjects ranging from 7 to 80 years old and used the data to develop a deep learning end-to-end model ShoeNet to analyze age-related patterns and predict age. The model integrates various convolutional neural network models together using a skip mechanism to extract age-related features, especially in pressure and abrasion regions from pair-wise shoeprints. The results show that 40.23% of the subjects had prediction errors within 5-years of age and the prediction accuracy for gender classification reached 86.07%. Interestingly, the age-related features mostly reside in the asymmetric differences between left and right shoeprints. The analysis also reveals interesting age-related and gender-related patterns in the pressure distributions on shoeprints; in particular, the pressure forces spread from the middle of the toe toward outside regions over age with gender-specific variations on heel regions. Such statistics provide insight into new methods for forensic investigations, medical studies of gait-pattern disorders, biometrics, and sport studies.

cs.CV↗

arXiv2019

Lunar impact craters identification and age estimation with Chang'E data by deep and transfer learning

Impact craters, as "lunar fossils", are the most dominant lunar surface features and occupy most of the Moon's surface. Their formation and evolution record the history of the Solar System. Sixty years of triumphs in the lunar exploration projects accumulated a large amount of lunar data. Currently, there are 9137 existing recognized craters. However, only 1675 of them have been determined age, which is obviously not satisfactory to reveal the evolution of the Moon. Identifying craters is a challenging task due to their enormous difference in size, large variations in shape and vast presence. Furthermore, estimating the age of craters is extraordinarily difficult due to their complex and different morphologies. Here, in order to effectively identify craters and estimate their age, we convert the crater identification problem into a target detection task and crater age estimation into a taxonomy structure. From an initial small number of available craters, we progressively identify craters and estimate their age from Chang'E data by transfer learning (TL) using deep neural networks. For comprehensive identification of multi-scale craters, a two-stage craters detection approach is developed. Thus 117240 unrecognized lunar craters that range in diameter from 532 km to 1 km are identified. Then, a two-stage classification approach is developed to estimate the age of craters by simultaneously extracting their morphological features and stratigraphic information. The age of 79243 craters larger than 3 km in diameter is estimated. These identified and aged craters throughout the mid and low-latitude regions of the Moon are crucial for reconstructing the dynamic evolution process of the Solar System.

astro-ph.EP↗

arXiv2018

Formation of the Abundance Boundaries of the Heavier Neutron-capture Elements in Metal-poor Stars

The abundance scatter of heavier r-process elements (Z > 56) relative to Fe ([r/Fe]) in metal-poor stars preserves excellent information of the star formation history and provides important insights into the various situations of the Galactic chemical enrichment. In this respect, the upper and lower boundaries of [r/Fe] could present useful clues for investigating the extreme situations of the star formation history and the early Galactic chemical evolution. In this paper, we investigate the formation of the upper and lower boundaries of [r/Fe] for the gas clouds. We find that, for a cloud from which metal-poor stars formed, the formation of the upper limits of [r/Fe] is mainly due to the pollution from a single main r-process event. For a cloud from which metal-poor stars formed, the formation of the lower limits of [r/Fe] is mainly due to the pollution from a single SN II event that ejects primary Fe.

astro-ph.SR↗

arXiv2017

Compositional Learning of Relation Path Embedding for Knowledge Base Completion

Large-scale knowledge bases have currently reached impressive sizes; however, these knowledge bases are still far from complete. In addition, most of the existing methods for knowledge base completion only consider the direct links between entities, ignoring the vital impact of the consistent semantics of relation paths. In this paper, we study the problem of how to better embed entities and relations of knowledge bases into different low-dimensional spaces by taking full advantage of the additional semantics of relation paths, and we propose a compositional learning model of relation path embedding (RPE). Specifically, with the corresponding relation and path projections, RPE can simultaneously embed each entity into two types of latent spaces. It is also proposed that type constraints could be extended from traditional relation-specific constraints to the new proposed path-specific constraints. The results of experiments show that the proposed model achieves significant and consistent improvements compared with the state-of-the-art algorithms.

cs.CL↗

arXiv2016

Footprints of the weak s-process in the carbon-enhanced metal-poor star ET0097

Historically, the weak s-process contribution to metal-poor stars is thought to be extremely small, due to the effect of the secondary-like nature of the neutron source 22Ne(a;n)25Mg in massive stars, which means that metal-poor weak s-process stars could not be found. ET0097 is the first observed carbon-enhanced metal-poor (CEMP) star in the Sculptor dwarf spheroidal galaxy. Because C is enriched and the elements heavier than Ba are not overabundant, ET0097 can be classified as a CEMP-no star. However, this star shows overabundances of lighter n-capture elements (i.e., Sr, Y and Zr). In this work, having adopted the abundance decomposition approach, we investigate the astrophysical origins of the elements in ET0097. We find that the light elements and iron-peak elements (from O to Zn) of the star mainly originate from the primary process of massive stars and the heavier n-capture elements (heavier than Ba) mainly come from the main r-process. However, the lighter n-capture elements such as Sr, Y and Zr should mainly come from the primary weak s-process. The contributed fractions of the primary weak s-process to the Sr, Y and Zr abundances of ET0097 are about 82%, 84% and 58% respectively, suggesting that the CEMP star ET0097 should have the footprints of the weak s-process. The derived result should be a significant evidence that the weak s-process elements can be produced in metal-poor massive stars.

astro-ph.SR↗

arXiv2013

Modelling narrow-line regions of active galaxies in the Sloan Digital Sky Survey - I.Sample selection and physical conditions

We systematically determine the electron density and electron temperature of active galaxies and star-forming galaxies using spectroscopy from the SDSS DR7, while mainly focusing on the NLRs. Density and temperature are determined through the I[S II] 6716/6731 and I[O III] 5007/4363 ratios, respectively, in our [O III] 4363 emission sample of 15 019 galaxies. We find two sequences and the typical range of density in the NLRs of AGNs is 100-1000 /cm3. The temperatures in the NLRs range from 10 000 to 20 000 K for Seyferts, and the ranges were even higher and wider for LINERs and composites. We also propose that Y_LINER ~ Y_Seyfert > Y_composite > Y_star-forming, where Y is the characteristic present-day star-formation time-scale. While in the AGN case, we find several strong lines of evidence suggest that some supplementary energy source(s) should be responsible for high ionization potential.

astro-ph.CO↗

arXiv2011

Automatic Determination of Stellar Atmospheric Parameters and Construction of Stellar Spectral Templates of the Guoshoujing Telescope (LAMOST)

A number of spectroscopic surveys have been carried out or are planned to study the origin of the Milky Way. Their exploitation requires reliable automated methods and softwares to measure the fundamental parameters of the stars. Adopting the ULySS package, we have tested the effect of different resolutions and signal-to-noise ratios (SNR) on the measurement of the stellar atmospheric parameters (effective temperature Teff, surface gravity log g, and metallicity [Fe/H]). We show that ULySS is reliable to determine these parameters with medium-resolution spectra (R~2000). Then, we applied the method to measure the parameters of 771 stars selected in the commissioning database of the Guoshoujing Telescope (GSJT). The results were compared with the SDSS/SEGUE Stellar Parameter Pipeline (SSPP), and we derived precisions of 167 K, 0.34 dex, and 0.16 dex for Teff, log g and [Fe/H] respectively. Furthermore, 120 of these stars are selected to construct the primary stellar spectra template library (Version 1.0) of GSJT, and will be deployed as basic ingredients for the GSJT automated parametrization pipeline.

astro-ph.SR↗