Large language models (LLMs) enhanced with retrieval-augmented generation (RAG) techniques still face difficulties delivering trustworthy question answering (QA) in specialized areas like radiation biology. This field is closely linked to human health and disease risks, making interpretable and accurate QA systems particularly important for scientific research and public understanding. To address these challenges, we present BioRadRAG (https://www.biosino.org/radiation/), a domain-specific RAG-based QA system designed to provide evidence-grounded answers in radiation biology. BioRadRAG constructs a curated knowledge base of$\mathbf{1 0, 6 6 9}$publications and adopts a parent-child chunking strategy to improve retrieval granularity and contextual coherence. We propose a multi-stage reasoning pipeline combining question classification, semantic rewriting, and reranking, all guided by structured prompts for response generation, thereby enhancing factual grounding and traceability. Two benchmark datasets were curated for evaluation. Experimental results demonstrate that BioRadRAG improves F1 scores by$\mathbf{7 \% - 1 1 \%}$on objective QA tasks compared to general-purpose LLMs and achieves better performance in subjective question evaluations. Ablation analysis further shows that the semantic rewriting and reranking modules contribute complementary gains in recall and precision, validating the multi-stage pipeline design. Furthermore, the system supports knowledge source filtering by publication year and journal impact factor, paragraph-level localization, mind map generation, and multi-turn interaction, enhancing accessibility and transparency in the exploration of scientific information. These findings highlight BioRadRAG's potential to advance domainspecific, explainable QA in radiation biology research.

Spoofing is a deceptive trading strategy where fraudsters place a large number of fake orders to manipulate market prices, severely distorting market fairness and threatening market stability.With the advancement of fraudulent tactics, spoofing patterns span across various levels of interaction, involving not only the local structure of individual spoofing transactions but also spoofing groups and global patterns.Relying solely on local context makes it challenging to capture multi-granularity risk signals, especially for organized and covert spoofing.Additionally, existing methods fail to consider the differences and relative importance between features of varying granularity, leading to feature distortion and noise.Therefore, we propose a multi-granularity augmented graph learning method that differentially captures fraud signals at local, group, and global levels.It utilizes multi-hop differential aggregation and communityaugmented strategy to capture information from local to global perspectives, adaptively distinguishing the contributions of different granularity.To avoid excessive fusion of multi-granularity information, we combine contrastive loss and cross-entropy loss for joint optimization, preserving key features while enhancing the method's robustness and accuracy.Extensive experiments on real-world datasets demonstrate the effectiveness of our proposed approach in spoofing detection, providing a robust solution for regulatory agencies.Our work will help financial institutions enhance their regulatory capabilities, protect investors' interests, and promote the healthy development of financial markets.

Spoofing detection in financial trading is crucial, especially for identifying complex behaviors such as conspiracy spoofing. Traditional machine-learning approaches primarily focus on isolated node features, often overlooking the broader context of interconnected nodes. Graph-based techniques, particularly Graph Neural Networks (GNNs), have advanced the field by leveraging relational information effectively. However, in real-world spoofing detection datasets, trading behaviors exhibit dynamic, irregular patterns. Existing spoofing detection methods, though effective in some scenarios, struggle to capture the complexity of dynamic and diverse, evolving inter-node relationships. To address these challenges, we propose a novel framework called the Generative Dynamic Graph Model (GDGM), which models dynamic trading behaviors and the relationships among nodes to learn representations for conspiracy spoofing detection. Specifically, our approach incorporates the generative dynamic latent space to capture the temporal patterns and evolving market conditions. Raw trading data is first converted into time-stamped sequences. Then we model trading behaviors using the neural ordinary differential equations and gated recurrent units, to generate the representation incorporating temporal dynamics of spoofing patterns. Furthermore, pseudo-label generation and heterogeneous aggregation techniques are employed to gather relevant information and enhance the detection performance for conspiratorial spoofing behaviors. Experiments conducted on spoofing detection datasets demonstrate that our approach outperforms state-of-the-art models in detection accuracy. Additionally, our spoofing detection system has been successfully deployed in one of the largest global trading markets, further validating the practical applicability and performance of the proposed method.

Fact verification constitutes a pivotal application in the effort to combat the dissemination of disinformation, a concern that has recently garnered considerable attention. However, previous studies in the field of fact verification, particularly those focused on question-answering dialogue, have exhibited limitations, such as failing to fully exploit the potential of question structures and ignoring relevant label information during the verification process. In this paper, we introduce Label-Infused Iterative Information Interacting (LI4), a novel approach designed for the task of question-answering dialogue based fact verification. LI4 consists of two meticulously designed components, namely the Iterative Information Refining and Filtering Module (IIRF) and the Fact Label Embedding Module (FLEM). The IIRF uses the Interactive Gating Mechanism to iteratively filter out the noise of question and evidence, concurrently refining the claim information. The FLEM is conceived to strengthen the understanding ability of the model towards labels by injecting label knowledge. We evaluate the performance of the proposed LI4 on HEALTHVER, FAVIQ, and COLLOQUIAL. The experimental results confirm that our LI4 model attains remarkable progress, manifesting as a new state-of-the-art performance.

Stateful policies play an important role in reinforcement learning, such as handling partially observable environments, enhancing robustness, or imposing an inductive bias directly into the policy structure. The conventional method for training stateful policies is Backpropagation Through Time (BPTT), which comes with significant drawbacks, such as slow training due to sequential gradient propagation and the occurrence of vanishing or exploding gradients. The gradient is often truncated to address these issues, resulting in a biased policy update. We present a novel approach for training stateful policies by decomposing the latter into a stochastic internal state kernel and a stateless policy, jointly optimized by following the stateful policy gradient. We introduce different versions of the stateful policy gradient theorem, enabling us to easily instantiate stateful variants of popular reinforcement learning and imitation learning algorithms. Furthermore, we provide a theoretical analysis of our new gradient estimator and compare it with BPTT. We evaluate our approach on complex continuous control tasks, e.g. humanoid locomotion, and demonstrate that our gradient estimator scales effectively with task complexity while offering a faster and simpler alternative to BPTT.

The phenomenon of related trading, in which organized communities engage in coordinated trading activities, poses a significant threat to the financial markets. Such activities can facilitate financial fraud, such as insider trading, price control, and market manipulation. Therefore, the detection of related trading is crucial for regulators to take actions to maintain market fairness and reduce market risk. The key challenge in detecting related trading is to measure the relevance of the trading behaviors of traders. Trade data is often represented as time series data, and traditional methods for analyzing such data typically focus on designing complex handcrafted features to measure the similarity of these time series. However, these methods are often incapable of capturing the complexity and variability of trading behaviors, which can be confusing and misleading. In this paper, we address this limitation by introducing a Hierarchical Transformer-based Siamese Network (HTSN) for related trading detection. The HTSN learns the correlation of the trade data from two accounts in an end-to-end manner, and is able to better capture trading information by splitting the data into different scales and stacking multiple Transformer encoders to extract features hierarchically. The experimental results on real-world data from China's futures market, indicate that the proposed HTSN model substantially outperforms previous approaches in detecting related trading.

We propose a video feature representation learning frame-work called STAR-GNN, which applies a pluggable graph neural network component on a multi-scale lattice feature graph. The essence of STAR-GNN is to exploit both the temporal dynamics and spatial contents as well as vi-sual connections between regions at different scales in the frames. It models a video with a lattice feature graph in which the nodes represent regions of different granularity, with weighted edges that represent the spatial and temporal links. The contextual nodes are aggregated simultaneously by graph neural networks with parameters trained with re-trieval triplet loss. In the experiments, we show that STAR-GNN effectively implements a dynamic attention mechanism on video frame sequences, resulting in the emphasis for dy-namic and semantically rich content in the video, and is robust to noise and redundancies. Empirical results show that STAR-GNN achieves state-of-the-art performance for Content-Based Video Retrieval.

Convolution neural network (CNN) has been successfully applied to hyperspectral image classification. However, multiclass imbalance is a major problem in the classification of hyper spectral images, and traditional CNN can hardly improve the accuracy of minority classes effectively. In this paper, a new ensemble CNN with enhanced feature subspaces (ECNN-EFSs) algorithm is proposed, which utilizes an imbalanced training set to train the model and achieves accurate classification. Experimental results on two common hyperspectral datasets show that the proposed algorithm outperforms the traditional CNN and ensemble CNN algorithms.

Convolutional neural networks (CNN) can automatically learn features from the hyperspectral image data, which could avoid the difficulty of manually extracting features. However, the number of training set for the classification of hyperspectral images is always limited, making it difficult for CNN to obtain effective features and resulting in low classification accuracy. In this paper, a spectral-spatial feature (SSF) extraction based CNN method is proposed for an accurate classification with a small training set. Experimental results based on two standard hyperspectral images demonstrate the effectiveness of the proposed method.

We bring the novel idea of exploiting motifs into network embedding, in a dual-level network representation learning model called RUM (network Representation learning Using Motifs). Towards the leveraging of graph motifs that constitute higher-order organizations in a network, we propose two strategies, namely MotifWalk and MotifRe-weighting for learning motif-aware network embeddings. Motif-based and node-based representations are simultaneously generated, so that both the high-order structures and each node's individual properties are preserved in the final embeddings. We demonstrate that RUM has strong and well-balanced capability of preserving lowerorder proximities while discovering and capturing higher-order network structures. In empirical evaluation, RUM is tested on multiple public datasets, that range from small to medium citation networks to a large social network with more than a million nodes. Results show that the use of motifs in the representation learning process brings substantial benefits in reallife tasks, resulting in up to 12% microF1 and 8% macroF1 relative gains for node classification performance over the bestperforming competing methods.

The denoising and deblurring of images are the two essential restoration tasks in the document image processing task. As the preprocessing stages of the processing pipeline, the quality of denoising and deblurring heavily influences the result of subsequent tasks, such as character detection and recognition. In this paper, we propose a novel neural method for restoring document images. We named our network Skip-Connected Deep Convolutional Autoencoder (SCDCA), which is composed of multiple layers of convolution followed by a batch normalization layer and the leaky rectified linear unit (Leaky ReLU) activation function. Inspired by the idea of residual learning, we use two types of skip connections in the network. One is identity mapping between convolution layers and the other is used to connect the input and output. Through these connections, the network learns the residual between the noisy and clean images instead of learning an ordinary transformation function. We empirically evaluate our algorithm on an open and challenging document images dataset. We also assess our restoring results using the optical character recognition (OCR) test. Experimental results have demonstrated the effectiveness and efficiency of our proposed algorithm by comparing with several state-of-the-art methods.

Recently, brain-inspired machine intelligence has gained great attention, research indicates that human brain grows in a self-organizing manner, especially in the visual cortex, fast object recognition is performed through distinct structures of neural connections and receptive fields. It is widely believed that such inhomogeneity is evolved through self-organizing by a process of neural plasticity. In this paper, a hierarchical self-organization spiking neural network (SOSNN) is proposed to solve the object recognition task with reinforcement plasticity rule, which simulates human visual cortex incorporating with many neural mechanisms like synaptic plasticity, homeostasis plasticity and lateral inhibitory. There are two phases in SOSNN which conform to the human visual pathway, feature extraction and decision-making (recognition). The object recognition task is performed in a manner of “end-to-end” in SOSNN since the network's decision is made by spiking activities in last layer without any external classifier. SOSNN is trained with reward-modulated spiking-time-dependent-plasticity (RM-STDP) rule in a reinforcement form, and the unsupervised STDP rule is also used to compare with RM-STDP. The classification experimental results on CIFAR and MNIST datasets show that SOSNN equipped with RM-STDP learning outperforms STDP and other existing unsupervised SNNs, which indicates the superiority of the proposed SOSNN in object recognition, and also testifies the reinforcement learning is an effective way to improve the performance of SNNs.

Person re-identification(ReID) is a task of associating persons that cross the non-overlapping camera views at different locations and times. It is a challenging task due to the large variations in person pose, background, luminance, occlusion, low resolution, etc. How to extracting a powerful features representation is the prime problem in ReID and is still unsolved. In this paper, we propose a cascade network architecture combined with a generative adversarial networks(GANs) and a convolutional neural network(CNN) to improve the performance of person re-identification. The GANs first generates the person body parts segmentation from the person image, and then inputs the segmentation label into the connected CNN together with the original person image. Finally obtain a discriminative and robust feature representation for ReID task. The body parts segmentation partitioning the person image into multiple segments, such as background, head, face, arms, lags, etc. The body parts segmentation information contains accurate borders and category attributes for body parts, which makes the our model more accurate compared to other predefined rigid parts alignment models. Experiments are conduced on the CUHK03, Market1501, DukeMTMC-ReID datasets and the results demonstrate that this approach outperforms several existing state-of-the-art methods.

UNLABELLED: Investigation of transcription factors (TFs) is of extreme significance for gleaning more information about the mechanisms underlying the dynamic transcriptional regulatory network. Herein, proTF is constructed to serve as a comprehensive data resource and phylogenomics analysis platform for prokaryotic TFs. It has many prominent characteristics: (i) detailed annotation information, including basic sequence features, domain organization, sequence homolog and sequence composition, was extensively collected, and then visually displayed for each TF entry in all prokaryotic genomes; (ii) workset was employed as the basic frame to provide an efficient way to organize the retrieved data and save intermediate records; and (iii) a number of elaborated tools for phylogenomics analysis were implemented to investigate the evolutionary roles of specific TFs. In conclusion, proTF dedicates to the prokaryotic TFs with integrated multi-function, which will become a valuable resource for prokaryotic transcriptional regulatory network in the post-genomic era. AVAILABILITY: http://centre.bioinformatics.zj.cn/proTF.