❌

Normal view

Received — 5 February 2026 ⏭ Favourites

From Data to Behavior: Predicting Unintended Model Behaviors Before Training

✇cs.AI, q-bio.NC updates on arXiv.org
By: Mengru Wang · Zhenqian Xu · Junfeng Fang · Yunzhi Yao · Shumin Deng · Huajun Chen · Ningyu Zhang
5 February 2026 at 13:00
arXiv:2602.04735v1 Announce Type: cross Abstract: Large Language Models (LLMs) can acquire unintended biases from seemingly benign training data even without explicit cues or malicious content. Existing methods struggle to detect such risks before fine-tuning, making post hoc evaluation costly and inefficient. To address this challenge, we introduce Data2Behavior, a new task for predicting unintended model behaviors prior to training. We also propose Manipulating Data Features (MDF), a lightweight approach that summarizes candidate data through their mean representations and injects them into the forward pass of a base model, allowing latent statistical signals in the data to shape model activations and reveal potential biases and safety risks without updating any parameters. MDF achieves reliable prediction while consuming only about 20% of the GPU resources required for fine-tuning. Experiments on Qwen3-14B, Qwen2.5-32B-Instruct, and Gemma-3-12b-it confirm that MDF can anticipate unintended behaviors and provide insight into pre-training vulnerabilities.
Received — 14 January 2026 ⏭ Favourites

Moral Lenses, Political Coordinates: Towards Ideological Positioning of Morally Conditioned LLMs

✇cs.AI, q-bio.NC updates on arXiv.org
By: Chenchen Yuan · Bolei Ma · Zheyu Zhang · Bardh Prenkaj · Frauke Kreuter · Gjergji Kasneci
14 January 2026 at 13:00
arXiv:2601.08634v1 Announce Type: cross Abstract: While recent research has systematically documented political orientation in large language models (LLMs), existing evaluations rely primarily on direct probing or demographic persona engineering to surface ideological biases. In social psychology, however, political ideology is also understood as a downstream consequence of fundamental moral intuitions. In this work, we investigate the causal relationship between moral values and political positioning by treating moral orientation as a controllable condition. Rather than simply assigning a demographic persona, we condition models to endorse or reject specific moral values and evaluate the resulting shifts on their political orientations, using the Political Compass Test. By treating moral values as lenses, we observe how moral conditioning actively steers model trajectories across economic and social dimensions. Our findings show that such conditioning induces pronounced, value-specific shifts in models' political coordinates. We further notice that these effects are systematically modulated by role framing and model scale, and are robust across alternative assessment instruments instantiating the same moral value. This highlights that effective alignment requires anchoring political assessments within the context of broader social values including morality, paving the way for more socially grounded alignment techniques.
Received — 7 January 2026 ⏭ Favourites

The role of PCMT1 in prognosis tumor immune microenvironment and therapeutic responses across cancers

✇Omics In Lung
By: Bo Wang · Sijia Huang · Ruizhen Ren · Ruiqian Yao · Erwen Kou · Haixia Zhao · Hao Zhu · Mengyu Zhang · Liangzhe Wang · Yuanjie Zhu
6 January 2026 at 19:00

Discov Oncol. 2026 Jan 5. doi: 10.1007/s12672-025-04366-2. Online ahead of print.

ABSTRACT

BACKGROUND: Emerging evidence highlights the overexpression of Protein-L-isoaspartate (D-aspartate) O-methyltransferase (PCMT1) in multiple malignancies. However, its pan-cancer prognostic significance, tumor immune microenvironment (TIME) interactions, and therapeutic implications remain underexplored.

METHODS: Multi-omics data were integrated from UCSC Xena, GTEx, UALCAN, and published cohorts. PCMT1 expression patterns were systematically analyzed across 33 cancer types. Associations between PCMT1 and clinical outcomes, immune infiltration, immune checkpoint genes (ICGs), tumor mutation burden (TMB), microsatellite instability (MSI), and drug sensitivity were evaluated using bioinformatics pipelines.

RESULTS: Our pan-cancer analysis revealed differential expression patterns of PCMT1 across various malignancies, with significant upregulation in 20 cancer types and downregulation in 3 cancer types. Notably, PCMT1 overexpression was predominantly observed in epithelial-origin tumors, such as ACC (adrenocortical carcinoma), BRCA (breast invasive carcinoma), COAD (colon adenocarcinoma), and LUAD (lung adenocarcinoma). Survival analysis demonstrated that elevated PCMT1 expression was significantly correlated with unfavorable prognosis in multiple epithelial tumors, particularly in BRCA, esophageal carcinoma (ESCA), head and neck squamous cell carcinoma (HNSC), liver hepatocellular carcinoma (LIHC), and mesothelioma (MESO). Furthermore, comprehensive analysis identified significant associations between PCMT1 expression and various tumor microenvironment features, including immune scores, six distinct immune cell types, four immunosuppressive cell populations, cancer-associated fibroblasts (CAFs)-related markers, and immunosuppressive factors. PCMT1 expression also showed significant correlations with tumor mutation burden (TMB), microsatellite instability (MSI), DNA stemness score (DNAss), and RNA stemness score (RNAss). Particularly noteworthy was the strong positive correlation between PCMT1 expression and CAFs infiltration, along with their associated factors. These findings were further validated in independent immunotherapy cohorts, where PCMT1 consistently demonstrated immunosuppressive characteristics.

CONCLUSION: Multi-omics analysis suggests that PCMT1 may serve as a potential prognostic biomarker and a novel immunotherapy target for pan-cancer.

PMID:41491065 | DOI:10.1007/s12672-025-04366-2

Digital Twin AI: Opportunities and Challenges from Large Language Models to World Models

✇cs.AI, q-bio.NC updates on arXiv.org
By: Rong Zhou · Dongping Chen · Zihan Jia · Yao Su · Yixin Liu · Yiwen Lu · Dongwei Shi · Yue Huang · Tianyang Xu · Yi Pan · Xinliang Li · Yohannes Abate · Qingyu Chen · Zhengzhong Tu · Yu Yang · Yu Zhang · Qingsong Wen · Gengchen Mai · Sunyang Fu · Jiachen Li · Xuyu Wang · Ziran Wang · Jing Huang · Tianming Liu · Yong Chen · Lichao Sun · Lifang He
6 January 2026 at 13:00
arXiv:2601.01321v1 Announce Type: new Abstract: Digital twins, as precise digital representations of physical systems, have evolved from passive simulation tools into intelligent and autonomous entities through the integration of artificial intelligence technologies. This paper presents a unified four-stage framework that systematically characterizes AI integration across the digital twin lifecycle, spanning modeling, mirroring, intervention, and autonomous management. By synthesizing existing technologies and practices, we distill a unified four-stage framework that systematically characterizes how AI methodologies are embedded across the digital twin lifecycle: (1) modeling the physical twin through physics-based and physics-informed AI approaches, (2) mirroring the physical system into a digital twin with real-time synchronization, (3) intervening in the physical twin through predictive modeling, anomaly detection, and optimization strategies, and (4) achieving autonomous management through large language models, foundation models, and intelligent agents. We analyze the synergy between physics-based modeling and data-driven learning, highlighting the shift from traditional numerical solvers to physics-informed and foundation models for physical systems. Furthermore, we examine how generative AI technologies, including large language models and generative world models, transform digital twins into proactive and self-improving cognitive systems capable of reasoning, communication, and creative scenario generation. Through a cross-domain review spanning eleven application domains, including healthcare, aerospace, smart manufacturing, robotics, and smart cities, we identify common challenges related to scalability, explainability, and trustworthiness, and outline directions for responsible AI-driven digital twin systems.
Received — 16 December 2025 ⏭ Favourites

From User Interface to Agent Interface: Efficiency Optimization of UI Representations for LLM Agents

✇cs.AI, q-bio.NC updates on arXiv.org
By: Dezhi Ran · Zhi Gong · Yuzhe Guo · Mengzhou Wu · Yuan Cao · Haochuan Lu · Hengyu Zhang · Xia Zeng · Gang Cao · Liangchao Yao · Yuetang Deng · Wei Yang · Tao Xie
16 December 2025 at 13:00
arXiv:2512.13438v1 Announce Type: cross Abstract: While Large Language Model (LLM) agents show great potential for automated UI navigation such as automated UI testing and AI assistants, their efficiency has been largely overlooked. Our motivating study reveals that inefficient UI representation creates a critical performance bottleneck. However, UI representation optimization, formulated as the task of automatically generating programs that transform UI representations, faces two unique challenges. First, the lack of Boolean oracles, which traditional program synthesis uses to decisively validate semantic correctness, poses a fundamental challenge to co-optimization of token efficiency and completeness. Second, the need to process large, complex UI trees as input while generating long, compositional transformation programs, making the search space vast and error-prone. Toward addressing the preceding limitations, we present UIFormer, the first automated optimization framework that synthesizes UI transformation programs by conducting constraint-based optimization with structured decomposition of the complex synthesis task. First, UIFormer restricts the program space using a domain-specific language (DSL) that captures UI-specific operations. Second, UIFormer conducts LLM-based iterative refinement with correctness and efficiency rewards, providing guidance for achieving the efficiency-completeness co-optimization. UIFormer operates as a lightweight plugin that applies transformation programs for seamless integration with existing LLM agents, requiring minimal modifications to their core logic. Evaluations across three UI navigation benchmarks spanning Android and Web platforms with five LLMs demonstrate that UIFormer achieves 48.7% to 55.8% token reduction with minimal runtime overhead while maintaining or improving agent performance. Real-world industry deployment at WeChat further validates the practical impact of UIFormer.
Received — 3 November 2025 ⏭ Favourites

VeriMoA: A Mixture-of-Agents Framework for Spec-to-HDL Generation

✇cs.AI, q-bio.NC updates on arXiv.org
By: Heng Ping · Arijit Bhattacharjee · Peiyu Zhang · Shixuan Li · Wei Yang · Anzhe Cheng · Xiaole Zhang · Jesse Thomason · Ali Jannesari · Nesreen Ahmed · Paul Bogdan
3 November 2025 at 13:00
arXiv:2510.27617v1 Announce Type: new Abstract: Automation of Register Transfer Level (RTL) design can help developers meet increasing computational demands. Large Language Models (LLMs) show promise for Hardware Description Language (HDL) generation, but face challenges due to limited parametric knowledge and domain-specific constraints. While prompt engineering and fine-tuning have limitations in knowledge coverage and training costs, multi-agent architectures offer a training-free paradigm to enhance reasoning through collaborative generation. However, current multi-agent approaches suffer from two critical deficiencies: susceptibility to noise propagation and constrained reasoning space exploration. We propose VeriMoA, a training-free mixture-of-agents (MoA) framework with two synergistic innovations. First, a quality-guided caching mechanism to maintain all intermediate HDL outputs and enables quality-based ranking and selection across the entire generation process, encouraging knowledge accumulation over layers of reasoning. Second, a multi-path generation strategy that leverages C++ and Python as intermediate representations, decomposing specification-to-HDL translation into two-stage processes that exploit LLM fluency in high-resource languages while promoting solution diversity. Comprehensive experiments on VerilogEval 2.0 and RTLLM 2.0 benchmarks demonstrate that VeriMoA achieves 15--30% improvements in Pass@1 across diverse LLM backbones, especially enabling smaller models to match larger models and fine-tuned alternatives without requiring costly training.
Received — 9 October 2025 ⏭ Favourites

Identification of C4BPA as a genetically informed drug target in NSCLC: an integrative single-cell and multi-omics study based on the druggable genes

✇Omics In Lung
By: Zhihan Xiao · Xinji Liu · Wei Tang · Yan Lv · Tongyu Zhang · Xu Zhan · Qihang Sun · Willis Wasonga Omindo · Qi Wang · Ruijie Zhang · Wei Ping · Ni Zhang
7 October 2025 at 18:00

Hum Genomics. 2025 Oct 6;19(1):113. doi: 10.1186/s40246-025-00829-3.

ABSTRACT

BACKGROUND: Non-small cell lung cancer (NSCLC) remains a leading cause of cancer-related mortality worldwide. Despite advancements in treatment, drug resistance and limited therapeutic efficacy persist, underscoring the urgent need for novel and mechanistically informed therapeutic strategies. Identifying genetically supported drug targets may accelerate the development of precision therapies in NSCLC.

METHODS: We implemented an integrative multi-omics framework combining single-cell RNA sequencing (scRNA-seq), genome-wide association studies (GWAS), and molecular quantitative trait locus (QTL) datasets including expression (eQTL), protein (pQTL), and DNA methylation (mQTL) QTLs. Druggable candidates were systematically evaluated using a suite of Mendelian randomization (MR) approaches-including summary data-based MR (SMR), generalized SMR (GSMR), and genetic risk score (GRS) analysis. Epigenetic regulation and downstream signaling were further explored through mediation MR analysis.

RESULTS: C4BPA, a complement-regulatory macromolecule, emerged as a risk factor for NSCLC across multiple MR models, with consistent findings validated at both transcriptomic and proteomic levels. Epigenetic activation of C4BPA via DNA methylation was observed, and C4BPA expression was shown to promote NSCLC progression through the inflammatory chemokine CCL8 signaling axis. Sensitivity analyses confirmed the robustness of association inference.

CONCLUSIONS: Our findings identify C4BPA as a genetically validated and biologically plausible therapeutic target for NSCLC. This study demonstrates the power of integrating single-cell transcriptomics with population-scale omics and association inference to uncover actionable targets, offering a scalable framework for advancing precision oncology in lung cancer.

PMID:41053817 | PMC:PMC12502296 | DOI:10.1186/s40246-025-00829-3

Received — 24 July 2025 ⏭ Favourites

A 23-gene multi-omics signature predicts prognosis and treatment response in non-small cell lung cancer

✇(Multiomics OR Omics) AND (Lung OR gastric OR Hepatocellular)
By: Yinxu Zhang · Siwang Wang · Xiaoyang Chen · Guangyu Zhang · Yuxi Wang · Xiaomei Liu
23 July 2025 at 18:00

Discov Oncol. 2025 Jul 23;16(1):1391. doi: 10.1007/s12672-025-03243-2.

ABSTRACT

We developed the first multi-omics prognostic signature integrating 19 programmed cell death (PCD) pathways and organelle functions (mitochondria, lysosomes, Golgi apparatus) to predict prognosis and immunotherapy response in non-small cell lung cancer (NSCLC). (2) Methods: By combining single-cell RNA-seq, bulk transcriptomics, and deep neural networks (DNN), we identified a 23-gene signature validated across four cohorts (AUC 0.696–0.812). Conducted MR analysis to explore causal links between signature genes and NSCLC incidence, providing biological insights. (3) Results: A prognostic signature was developed, including 23 prognostic genes related to 19 PCD patterns and three organelle functions. The signature demonstrated powerful performance in predicting NSCLC prognosis, immune in-filtration, and therapeutic response. Established DNN models showed high value in predicting risk score groupings of NSCLC. MR analysis for combined SNP information of the 23 prognostic genes suggested a link to the high incidence of NSCLC. Individual MR analysis showed that HIF1A and SQLE expression had a causal effect on NSCLC incidence. (4) Conclusion: This signature stratifies high-risk patients with immunosuppressive microenvironments and predicts enhanced sensitivity to gemcitabine and PD-1 inhibitors, offering a roadmap for personalized NSCLC management.

SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s12672-025-03243-2.

PMID:40699399 | PMC:PMC12287486 | DOI:10.1007/s12672-025-03243-2

Received — 31 March 2025 ⏭ Favourites
  • ✇MRD
  • Liquid biopsies in cancer Hang Yin · Manjie Zhang · Yu Zhang · Xuebing Zhang · Xia Zhang · Bin Zhang
    Mol Biomed. 2025 Mar 20;6(1):18. doi: 10.1186/s43556-025-00257-8.ABSTRACTCancer ranks among the most lethal diseases worldwide. Tissue biopsy is currently the primary method for the diagnosis and biological analysis of various solid tumors. However, this method has some disadvantages related to insufficient tissue specimen collection and intratumoral heterogeneity. Liquid biopsy is a noninvasive approach for identifying cancer-related biomarkers in peripheral blood, which allows for repetitive s
     
  • 🏷️
  • ↗️

Liquid biopsies in cancer

✇MRD
By: Hang Yin · Manjie Zhang · Yu Zhang · Xuebing Zhang · Xia Zhang · Bin Zhang
20 March 2025 at 18:00

Mol Biomed. 2025 Mar 20;6(1):18. doi: 10.1186/s43556-025-00257-8.

ABSTRACT

Cancer ranks among the most lethal diseases worldwide. Tissue biopsy is currently the primary method for the diagnosis and biological analysis of various solid tumors. However, this method has some disadvantages related to insufficient tissue specimen collection and intratumoral heterogeneity. Liquid biopsy is a noninvasive approach for identifying cancer-related biomarkers in peripheral blood, which allows for repetitive sampling across multiple time points. In the field of liquid biopsy, representative biomarkers include circulating tumor cells (CTCs), circulating tumor DNA (ctDNA), and exosomes. Many studies have evaluated the prognostic and predictive roles of CTCs and ctDNA in various solid tumors. Although these studies have limitations, the results of most studies appear to consistently demonstrate the correlations of high CTC counts and ctDNA mutations with lower survival rates in cancer patients. Similarly, a reduction in CTC counts throughout therapy may be a potential prognostic indicator related to treatment response in advanced cancer patients. Moreover, the biochemical characteristics of CTCs and ctDNA can provide information about tumor biology as well as resistance mechanisms against targeted therapy. This review discusses the current clinical applications of liquid biopsy in cancer patients, emphasizing its possible utility in outcome prediction and treatment decision-making.

PMID:40108089 | PMC:PMC11923355 | DOI:10.1186/s43556-025-00257-8

Received — 1 February 2025 ⏭ Favourites

Efficient discovery of robust prognostic biomarkers and signatures in solid tumors

✇(Multiomics OR Omics) AND (Pancreatic)
By: Zaoqu Liu · Jinhai Deng · Hui Xu · Long Liu · Yuyuan Zhang · Yuhao Ba · Zhengyu Zhang · Fuchu He · Linhai Xie
26 January 2025 at 19:00

Cancer Lett. 2025 Mar 31;613:217502. doi: 10.1016/j.canlet.2025.217502. Epub 2025 Jan 24.

ABSTRACT

Recent advancements in multi-omics and big-data technologies have facilitated the discovery of numerous cancer prognostic biomarkers and gene signatures. However, their clinical application remains limited due to poor reproducibility and insufficient independent validation. Despite the availability of high-quality datasets, achieving reliable biomarker identification across multiple cohorts continues to be a significant challenge. To address these issues, we developed a comprehensive platform, SurvivalML, designed to support the discovery and validation of prognostic biomarkers and gene signatures using large-scale and harmonized data from 21 cancer types. Through SurvivalML, we identified DCLRE1B as a novel prognostic biomarker for hepatocellular carcinoma, with experimental confirmation of its role in promoting tumor progression. Additionally, we developed the Chinese glioblastoma prognostic signature (CGPS) and its simplified version, SCGPS, a three-gene model. Both demonstrated superior predictive performance compared to other glioblastoma signatures in our in-house cohort and five independent Chinese datasets. The SCGPS model was further validated in 109 clinical samples using multiplex immunofluorescence, showing strong consistency with the original CGPS model. Overall, SurvivalML provides a robust platform for the identification and validation of prognostic biomarkers and gene signatures, offering a valuable resource for advancing cancer research and clinical application.

PMID:39864538 | DOI:10.1016/j.canlet.2025.217502

Received — 24 December 2024 ⏭ Favourites

ScRNA-seq of gastric cancer tissues reveals differences in the immune microenvironment of primary tumors and metastases

✇Oncogene - Issue - nature.com science feeds
By: Yu Dong · Keshu Hu · Jiayu Zhang · Mengxuan Zhu · Mengling Liu · Yitao Yuan · Xun Sun · Zhenghang Xu · Suyao Li · Yanjing Zhu · Chi Zhang · Pengfei Zhang · Tianshu Liu
30 March 2024 at 08:00

Oncogene, Published online: 30 March 2024; doi:10.1038/s41388-024-03012-5

ScRNA-seq of gastric cancer tissues reveals differences in the immune microenvironment of primary tumors and metastases
There are no more articles
❌