Hebei University of Engineering

Wikipedia serves as the world's largest and most popular online reference encyclopedia, rich in structured knowledge and authoritative citations. Recently, numerous works have leveraged large language models to automatically generate Wikipedia-like articles. However, existing approaches primarily focus on producing singular narrative-type content, overlooking higher information-density structured elements such as timeline and table. To address these limitations, we propose WikiMAG, a multi-agent guided framework for generating structured Wikipedia-like articles. This framework employs a collaborative multi-agent mechanism to orchestrate the creation process, featuring three synergistic core components: Progressive planner first constructs the coarse-grained outline framework and then annotate fine-grained types for outline units, encompassing narrative, timeline, and table formats; Reflective inspector dynamically curates high-quality references via multi-round interactive feedback, thereby enhancing the authority and relevance of citations; Versatile writer integrates fine-grained outline details and high-quality reference information to generate information-rich articles, incorporating the three annotated formats. We evaluate WikiMAG on two public datasets, FreshWiki and WikiGenBen, across outline, writing, and verifiability dimensions. Compared with the best baseline method, our method achieves an average improvement of 6.73 points and 4.39 points in Heading Soft Recall and the METEOR metric (a machine translation and text generation evaluation metric) respectively, and an average increase of 16.84 percentage points in Citation Rate.

Long-form books are among the most information-rich and structurally complex forms of written content, often exceeding 100,000 words. While recent methods have enabled basic long-text generation, they remain limited in two key aspects: the inability to generate ultra-long content at book scale, and the lack of mechanisms for integrating rich factual information. To address these limitations, we propose DeepWriter, a multi-agent collaborative framework that follows a structured planning-then-generation paradigm. It first constructs a detailed book outline with narrative arcs and chapter semantics, then incrementally generates content conditioned on retrieved knowledge and contextual signals. DeepWriter supports controllable generation of full-length books exceeding 100,000 words, enriched with citations, trivia and images. To support evaluation beyond surface-level fluency, we introduce DeepWriter-Bench, a bilingual benchmark of 18 annotated books designed to assess book-scale coherence, richness, and factual grounding. Additionally, we propose BookScore, a unified 100-point metric for quantifying book maturity. Experimental results show that DeepWriter achieves a state-of-the-art BookScore of 80.92, consistently outperforming strong baselines.

BackgroundThis study aimed to analyze the breast cancer (BC) microenvironment using single-cell RNA sequencing (scRNA-seq) data, develop a prognostic stratification model, and identify potential therapeutic targets and drugs.MethodsscRNA-seq, bulk transcriptomic data, and clinical information were obtained from EMBL-EBI (single cell data from 17 tumor samples), TCGA (1039 tumor samples), and GEO databases [GSE20685 (324 tumor samples), GSE42568 (104 tumor samples), and GSE88770 (117 tumor samples)]. scRNA-seq data were used to analyze the tumor microenvironment landscape of BC and retrieve subcluster-specific markers (SSMs). Subtypes of BC patients were defined based on SSMs and bulk transcriptomic data, and a prognostic risk model was constructed using machine learning. The characteristic therapeutic targets of high-risk patients were further identified and potential drugs were evaluated by molecular docking and cellular thermal shift assay. In vitro and in vivo experiments were conducted to explore the functions of the core target, 3-hydroxybutyrate dehydrogenase 1 (BDH1) and tretinoin, which modulated the malignant biological behaviors of BC.ResultsA total of 2004 SSMs were identified. Six prognostic cell sub-clusters were identified, and 16 prognostic genes were identified from the SSMs of these six cell sub-clusters. Least absolute shrinkage and selection operator (LASSO)-Cox algorithm further identified four core genes, and a risk model was constructed. The overall survival time of BC patients in the high-risk group was shorter than that of the low-risk group, and the risk score was a predictor of prognosis in BC patients in both training dataset and validation datasets (P<0.0001 in all datasets). Six potential drug targets were identified in high-risk patients, five of which were significantly highly expressed in BC, and BDH1 was associated with the overall survival of BC patients. Tretinoin showed good binding activity for all five targets. Depletion of BDH1 or tretinoin inhibits the malignant biological behaviors of BC cells in vitro, and tretinoin suppressed the malignant biological behaviors of BC cells in vitro and in vivo, and reversed by BDH1 overexpression.ConclusionThe prognostic stratification model is promising for evaluating the clinical outcomes of BC patients, and BDH1 is an important drug target for high-risk patients, and tretinoin represses the malignancy of BC cells via targeting BDH1.