An Agentic Framework for Rapid Deployment of Edge AI Solutions in Industry 5.0

By: Jorge Martinez-Gil · Mario Pichler · Nefeli Bountouni · Sotiris Koussouris · Marielena M\'arquez Barreiro · Sergio Gusmeroli

31 October 2025 at 12:00

arXiv:2510.25813v1 Announce Type: new Abstract: We present a novel framework for Industry 5.0 that simplifies the deployment of AI models on edge devices in various industrial settings. The design reduces latency and avoids external data transfer by enabling local inference and real-time processing. Our implementation is agent-based, which means that individual agents, whether human, algorithmic, or collaborative, are responsible for well-defined tasks, enabling flexibility and simplifying integration. Moreover, our framework supports modular integration and maintains low resource requirements. Preliminary evaluations concerning the food industry in real scenarios indicate improved deployment time and system adaptability performance. The source code is publicly available at https://github.com/AI-REDGIO-5-0/ci-component.

Identity Management for Agentic AI: The new frontier of authorization, authentication, and security for an AI agent world

31 October 2025 at 12:00

arXiv:2510.25819v1 Announce Type: cross Abstract: The rapid rise of AI agents presents urgent challenges in authentication, authorization, and identity management. Current agent-centric protocols (like MCP) highlight the demand for clarified best practices in authentication and authorization. Looking ahead, ambitions for highly autonomous agents raise complex long-term questions regarding scalable access control, agent-centric identities, AI workload differentiation, and delegated authority. This OpenID Foundation whitepaper is for stakeholders at the intersection of AI agents and access management. It outlines the resources already available for securing today's agents and presents a strategic agenda to address the foundational authentication, authorization, and identity problems pivotal for tomorrow's widespread autonomous systems.

Supervised Reinforcement Learning: From Expert Trajectories to Step-wise Reasoning

cs.AI, q-bio.NC updates on arXiv.org

By: Yihe Deng · I-Hung Hsu · Jun Yan · Zifeng Wang · Rujun Han · Gufeng Zhang · Yanfei Chen · Wei Wang · Tomas Pfister · Chen-Yu Lee

31 October 2025 at 12:00

arXiv:2510.25992v1 Announce Type: cross Abstract: Large Language Models (LLMs) often struggle with problems that require multi-step reasoning. For small-scale open-source models, Reinforcement Learning with Verifiable Rewards (RLVR) fails when correct solutions are rarely sampled even after many attempts, while Supervised Fine-Tuning (SFT) tends to overfit long demonstrations through rigid token-by-token imitation. To address this gap, we propose Supervised Reinforcement Learning (SRL), a framework that reformulates problem solving as generating a sequence of logical "actions". SRL trains the model to generate an internal reasoning monologue before committing to each action. It provides smoother rewards based on the similarity between the model's actions and expert actions extracted from the SFT dataset in a step-wise manner. This supervision offers richer learning signals even when all rollouts are incorrect, while encouraging flexible reasoning guided by expert demonstrations. As a result, SRL enables small models to learn challenging problems previously unlearnable by SFT or RLVR. Moreover, initializing training with SRL before refining with RLVR yields the strongest overall performance. Beyond reasoning benchmarks, SRL generalizes effectively to agentic software engineering tasks, establishing it as a robust and versatile training framework for reasoning-oriented LLMs.

The Quest for Reliable Metrics of Responsible AI

cs.AI, q-bio.NC updates on arXiv.org

By: Theresia Veronika Rampisela · Maria Maistro · Tuukka Ruotsalo · Christina Lioma

31 October 2025 at 12:00

arXiv:2510.26007v1 Announce Type: cross Abstract: The development of Artificial Intelligence (AI), including AI in Science (AIS), should be done following the principles of responsible AI. Progress in responsible AI is often quantified through evaluation metrics, yet there has been less work on assessing the robustness and reliability of the metrics themselves. We reflect on prior work that examines the robustness of fairness metrics for recommender systems as a type of AI application and summarise their key takeaways into a set of non-exhaustive guidelines for developing reliable metrics of responsible AI. Our guidelines apply to a broad spectrum of AI applications, including AIS.

Multi-Agent Evolve: LLM Self-Improve through Co-evolution

cs.AI, q-bio.NC updates on arXiv.org

By: Yixing Chen · Yiding Wang · Siqi Zhu · Haofei Yu · Tao Feng · Muhan Zhang · Mostofa Patwary · Jiaxuan You

31 October 2025 at 12:00

arXiv:2510.23595v3 Announce Type: replace Abstract: Reinforcement Learning (RL) has demonstrated significant potential in enhancing the reasoning capabilities of large language models (LLMs). However, the success of RL for LLMs heavily relies on human-curated datasets and verifiable rewards, which limit their scalability and generality. Recent Self-Play RL methods, inspired by the success of the paradigm in games and Go, aim to enhance LLM reasoning capabilities without human-annotated data. However, their methods primarily depend on a grounded environment for feedback (e.g., a Python interpreter or a game engine); extending them to general domains remains challenging. To address these challenges, we propose Multi-Agent Evolve (MAE), a framework that enables LLMs to self-evolve in solving diverse tasks, including mathematics, reasoning, and general knowledge Q&A. The core design of MAE is based on a triplet of interacting agents (Proposer, Solver, Judge) that are instantiated from a single LLM, and applies reinforcement learning to optimize their behaviors. The Proposer generates questions, the Solver attempts solutions, and the Judge evaluates both while co-evolving. Experiments on Qwen2.5-3B-Instruct demonstrate that MAE achieves an average improvement of 4.54% on multiple benchmarks. These results highlight MAE as a scalable, data-efficient method for enhancing the general reasoning abilities of LLMs with minimal reliance on human-curated supervision.

Chain-of-Scrutiny: Detecting Backdoor Attacks for Large Language Models

cs.AI, q-bio.NC updates on arXiv.org

By: Xi Li · Ruofan Mao · Yusen Zhang · Renze Lou · Chen Wu · Jiaqi Wang

31 October 2025 at 12:00

arXiv:2406.05948v4 Announce Type: replace-cross Abstract: Large Language Models (LLMs), especially those accessed via APIs, have demonstrated impressive capabilities across various domains. However, users without technical expertise often turn to (untrustworthy) third-party services, such as prompt engineering, to enhance their LLM experience, creating vulnerabilities to adversarial threats like backdoor attacks. Backdoor-compromised LLMs generate malicious outputs to users when inputs contain specific "triggers" set by attackers. Traditional defense strategies, originally designed for small-scale models, are impractical for API-accessible LLMs due to limited model access, high computational costs, and data requirements. To address these limitations, we propose Chain-of-Scrutiny (CoS) which leverages LLMs' unique reasoning abilities to mitigate backdoor attacks. It guides the LLM to generate reasoning steps for a given input and scrutinizes for consistency with the final output -- any inconsistencies indicating a potential attack. It is well-suited for the popular API-only LLM deployments, enabling detection at minimal cost and with little data. User-friendly and driven by natural language, it allows non-experts to perform the defense independently while maintaining transparency. We validate the effectiveness of CoS through extensive experiments on various tasks and LLMs, with results showing greater benefits for more powerful LLMs.

Vital Insight: Assisting Experts' Context-Driven Sensemaking of Multi-modal Personal Tracking Data Using Visualization and Human-In-The-Loop LLM

cs.AI, q-bio.NC updates on arXiv.org

By: Jiachen Li · Xiwen Li · Justin Steinberg · Akshat Choube · Bingsheng Yao · Xuhai Xu · Dakuo Wang · Elizabeth Mynatt · Varun Mishra

31 October 2025 at 12:00

arXiv:2410.14879v4 Announce Type: replace-cross Abstract: Passive tracking methods, such as phone and wearable sensing, have become dominant in monitoring human behaviors in modern ubiquitous computing studies. While there have been significant advances in machine-learning approaches to translate periods of raw sensor data to model momentary behaviors, (e.g., physical activity recognition), there still remains a significant gap in the translation of these sensing streams into meaningful, high-level, context-aware insights that are required for various applications (e.g., summarizing an individual's daily routine). To bridge this gap, experts often need to employ a context-driven sensemaking process in real-world studies to derive insights. This process often requires manual effort and can be challenging even for experienced researchers due to the complexity of human behaviors. We conducted three rounds of user studies with 21 experts to explore solutions to address challenges with sensemaking. We follow a human-centered design process to identify needs and design, iterate, build, and evaluate Vital Insight (VI), a novel, LLM-assisted, prototype system to enable human-in-the-loop inference (sensemaking) and visualizations of multi-modal passive sensing data from smartphones and wearables. Using the prototype as a technology probe, we observe experts' interactions with it and develop an expert sensemaking model that explains how experts move between direct data representations and AI-supported inferences to explore, question, and validate insights. Through this iterative process, we also synthesize and discuss a list of design implications for the design of future AI-augmented visualization systems to better assist experts' sensemaking processes in multi-modal health sensing data.

Epistemic Diversity and Knowledge Collapse in Large Language Models

cs.AI, q-bio.NC updates on arXiv.org

By: Dustin Wright · Sarah Masud · Jared Moore · Srishti Yadav · Maria Antoniak · Chan Young Park · Isabelle Augenstein

31 October 2025 at 12:00

arXiv:2510.04226v4 Announce Type: replace-cross Abstract: Large language models (LLMs) tend to generate lexically, semantically, and stylistically homogenous texts. This poses a risk of knowledge collapse, where homogenous LLMs mediate a shrinking in the range of accessible information over time. Existing works on homogenization are limited by a focus on closed-ended multiple-choice setups or fuzzy semantic features, and do not look at trends across time and cultural contexts. To overcome this, we present a new methodology to measure epistemic diversity, i.e., variation in real-world claims in LLM outputs, which we use to perform a broad empirical study of LLM knowledge collapse. We test 27 LLMs, 155 topics covering 12 countries, and 200 prompt variations sourced from real user chats. For the topics in our study, we show that while newer models tend to generate more diverse claims, nearly all models are less epistemically diverse than a basic web search. We find that model size has a negative impact on epistemic diversity, while retrieval-augmented generation (RAG) has a positive impact, though the improvement from RAG varies by the cultural context. Finally, compared to a traditional knowledge source (Wikipedia), we find that country-specific claims reflect the English language more than the local one, highlighting a gap in epistemic representation

Integrating Genomics into Multimodal EHR Foundation Models

cs.AI, q-bio.NC updates on arXiv.org

By: Jonathan Amar · Edward Liu · Alessandra Breschi · Liangliang Zhang · Pouya Kheradpour · Sylvia Li · Lisa Soleymani Lehmann · Alessandro Giulianelli · Matt Edwards · Yugang Jia · David Nola · Raghav Mani · Pankaj Vats · Jesse Tetreault · T. J. Chen · Cory Y. McLean

31 October 2025 at 12:00

arXiv:2510.23639v2 Announce Type: replace-cross Abstract: This paper introduces an innovative Electronic Health Record (EHR) foundation model that integrates Polygenic Risk Scores (PRS) as a foundational data modality, moving beyond traditional EHR-only approaches to build more holistic health profiles. Leveraging the extensive and diverse data from the All of Us (AoU) Research Program, this multimodal framework aims to learn complex relationships between clinical data and genetic predispositions. The methodology extends advancements in generative AI to the EHR foundation model space, enhancing predictive capabilities and interpretability. Evaluation on AoU data demonstrates the model's predictive value for the onset of various conditions, particularly Type 2 Diabetes (T2D), and illustrates the interplay between PRS and EHR data. The work also explores transfer learning for custom classification tasks, showcasing the architecture's versatility and efficiency. This approach is pivotal for unlocking new insights into disease prediction, proactive health management, risk stratification, and personalized treatment strategies, laying the groundwork for more personalized, equitable, and actionable real-world evidence generation in healthcare.

Nanomaterial-assisted immunodiagnostic profiling and therapeutic targeting of hepatocellular carcinoma: from molecular biomarkers to clinical applications

(Multiomics OR Omics) AND (Lung OR gastric OR Hepatocellular)

By: Jiping Luo · Jianzeng Ye · Kaipeng Huang · Ziyu Cheng · Liming Liu · Xianpeng Li

30 October 2025 at 18:00

Front Immunol. 2025 Oct 14;16:1668630. doi: 10.3389/fimmu.2025.1668630. eCollection 2025.

ABSTRACT

AIMS AND OBJECTIVES: This study aimed to identify immunologically relevant transcriptomic and proteomic biomarkers in hepatocellular carcinoma (HCC) and to characterize their B-cell epitopes for potential integration into nanomaterial-based biosensors and immunomodulatory platforms for early diagnosis and targeted therapy.

METHODS: We conducted a comprehensive multi-omics analysis by integrating transcriptomic (TCGA-LIHC) and proteomic data to identify differentially expressed genes (DEGs) in HCC. Protein-protein interaction networks and pathway enrichment were used to prioritize hub genes. Five candidate biomarkers, RFC2, HSP90AB1, YWHAZ, CYP2E1, and ADH4, were selected for qRT-PCR and serum ELISA validation in clinical cohorts comprising 85 HCC patients and 50 healthy controls. B-cell epitope prediction was performed using BepiPred 2.0 and validated through synthetic peptide-based ELISA in the same cohort to assess immunoreactivity. Diagnostic performance was evaluated using ROC curve analysis.

RESULTS: RFC2, HSP90AB1, and YWHAZ were significantly upregulated (|log2FC|>0.2) and showed high serological expression, whereas CYP2E1 and ADH4 were consistently downregulated. Predicted B-cell epitopes from RFC2, HSP90AB1, and YWHAZ exhibited strong immunoreactivity (AUC>0.84), indicating their diagnostic potential. Enrichment analysis revealed that upregulated DEGs were involved in cell cycle and mitotic progression, while downregulated genes were linked to immune suppression and metabolic dysfunction. These validated immunogenic epitopes offer promising anchors for nanomaterial-functionalized biosensors, such as gold nanoparticle-conjugated ELISA, graphene-based electrochemical platforms, and peptide-coated quantum dots, for ultrasensitive and multiplexed HCC detection.

CONCLUSION: By integrating transcriptomic and proteomic screening with epitope-level validation, we identified a novel panel of immunogenic biomarkers suitable for nanomaterial-enabled diagnostics in HCC. These findings support the translational potential of peptide-nano scaffold conjugates in developing minimally invasive, immune-responsive biosensing and therapeutic tools tailored for early-stage liver cancer management.

PMID:41164201 | PMC:PMC12558944 | DOI:10.3389/fimmu.2025.1668630

Toward governance of artificial intelligence in pediatric healthcare

npj Digital Medicine

By: Felix Richter · Emma Holmes · Florian Richter · Katherine Guttmann · Son Q. Duong · Sandeep Gangadharan · Eric E. Schadt · Hojjat Salmasian · Bruce D. Gelb · Benjamin S. Glicksberg

30 October 2025 at 08:00

npj Digital Medicine, Published online: 30 October 2025; doi:10.1038/s41746-025-02000-7

Toward governance of artificial intelligence in pediatric healthcare

Reading view