❌

Normal view

Received — 12 January 2026 ⏭ Favourites

The Evaluation Gap in Medicine, AI and LLMs: Navigating Elusive Ground Truth & Uncertainty via a Probabilistic Paradigm

✇cs.AI, q-bio.NC updates on arXiv.org
By: Aparna Elangovan · Lei Xu · Mahsa Elyasi · Ismail Akdulum · Mehmet Aksakal · Enes Gurun · Brian Hur · Saab Mansour · Ravid Shwartz Ziv · Karin Verspoor · Dan Roth
12 January 2026 at 13:00
arXiv:2601.05500v1 Announce Type: new Abstract: Benchmarking the relative capabilities of AI systems, including Large Language Models (LLMs) and Vision Models, typically ignores the impact of uncertainty in the underlying ground truth answers from experts. This ambiguity is particularly consequential in medicine where uncertainty is pervasive. In this paper, we introduce a probabilistic paradigm to theoretically explain how high certainty in ground truth answers is almost always necessary for even an expert to achieve high scores, whereas in datasets with high variation in ground truth answers there may be little difference between a random labeller and an expert. Therefore, ignoring uncertainty in ground truth evaluation data can result in the misleading conclusion that a non-expert has similar performance to that of an expert. Using the probabilistic paradigm, we thus bring forth the concepts of expected accuracy and expected F1 to estimate the score an expert human or system can achieve given ground truth answer variability. Our work leads to the recommendation that when establishing the capability of a system, results should be stratified by probability of the ground truth answer, typically measured by the agreement rate of ground truth experts. Stratification becomes critical when the overall performance drops below a threshold of 80%. Under stratified evaluation, performance comparison becomes more reliable in high certainty bins, mitigating the effect of the key confounding factor -- uncertainty.

Benchmarking LLM-based Agents for Single-cell Omics Analysis

✇cs.AI, q-bio.NC updates on arXiv.org
By: Yang Liu · Lu Zhou · Xiawei Du · Ruikun He · Rongbo Shen · Yixue Li
12 January 2026 at 13:00
arXiv:2508.13201v2 Announce Type: replace-cross Abstract: The surge in multimodal single-cell omics data exposes limitations in traditional, manually defined analysis workflows. AI agents offer a paradigm shift, enabling adaptive planning, executable code generation, traceable decisions, and real-time knowledge fusion. However, the lack of a comprehensive benchmark critically hinders progress. We introduce a novel benchmarking evaluation system to rigorously assess agent capabilities in single-cell omics analysis. This system comprises: a unified platform compatible with diverse agent frameworks and LLMs; multidimensional metrics assessing cognitive program synthesis, collaboration, execution efficiency, bioinformatics knowledge integration, and task completion quality; and 50 diverse real-world single-cell omics analysis tasks spanning multi-omics, species, and sequencing technologies. Our evaluation reveals that Grok-3-beta achieves state-of-the-art performance among tested agent frameworks. Multi-agent frameworks significantly enhance collaboration and execution efficiency over single-agent approaches through specialized role division. Attribution analyses of agent capabilities identify that high-quality code generation is crucial for task success, and self-reflection has the most significant overall impact, followed by retrieval-augmented generation (RAG) and planning. This work highlights persistent challenges in code generation, long-context handling, and context-aware knowledge retrieval, providing a critical empirical foundation and best practices for developing robust AI agents in computational biology.

Personalizing Treatment for Pancreatic Ductal Adenocarcinoma: The Emerging Role of Minimal Residual Disease in Perioperative Decision-Making

✇MRD
By: Charalampos Theocharopoulos · Nikolaos Machairas · Ioannis A Ziogas · Benedetto Mungo · Marco Del Chiaro · Georgios K Glatzounis · Richard Schulick · Georgios C Sotiropoulos
10 January 2026 at 19:00

Cancers (Basel). 2025 Dec 27;18(1):94. doi: 10.3390/cancers18010094.

ABSTRACT

Pancreatic ductal adenocarcinoma (PDAC) is a highly aggressive malignancy with poor long-term survival despite advances in surgical techniques, systemic therapies, and perioperative management. High rates of systemic recurrence following curative-intent resection suggest that many patients harbor minimal residual disease (MRD), microscopic tumor burden that persists postoperatively and remains undetectable by conventional diagnostic tools. Recent advances in liquid biopsy technologies, particularly circulating tumor DNA (ctDNA) analysis, alongside detailed characterization of the PDAC mutational landscape, offer a promising non-invasive approach for MRD detection. Emerging evidence indicates that MRD status can serve as a sensitive prognostic biomarker, identify patients at high risk of relapse, and guide personalized perioperative therapy, including optimization of adjuvant treatment. This review summarizes current knowledge on the biology and detection of MRD in PDAC, its implications for perioperative risk stratification and treatment decision-making, and discusses future directions for integrating MRD assessment into clinical practice to enable more precise, individualized patient management.

PMID:41514607 | PMC:PMC12784771 | DOI:10.3390/cancers18010094

Received — 10 January 2026 ⏭ Favourites

BCMA-directed mRNA CAR-T cell therapy for myasthenia gravis: exploratory biomarker analysis of a placebo-controlled phase 2b trial

✇Nature Medicine
By: Renee R. Fedak · Rachel N. Ruggerie · Yufei Shan · Elizabeth J. Curvino · Juliana F. de Sousa · Shaji Daniel · Minhtran Ngo-Casi · Hafsa Kamboh · Tuan Vu · Hacer Durmuş · Tahseen Mozaffar · James F. Howard Jr. · Emily P. English · Albina Benson · Matthew T. Duvernay · Michael S. Singer · Murat V. Kalayoglu · Carsten Brunn · Aaron Bodansky · Mark S. Anderson · Joseph L. DeRisi · Samantha T. Garcia · David J. L. Yu · Kelsey C. Zorn · Metin Kurtoglu · Miloš D. Miljković · C. Andrew Stewart · Christopher M. Jewell · Charito Buensuceso · Lisa H. Tostanoski · Yi Zhang · Heng Xu · Chad A. Pope · Dina Schneider · Adam English · Atanas Filev · Jennifer Biaksangi · George Small · Michael Rivner · Michael A. Badruddoja · Bennett Myers · Sheetal Shroff · Adam Slanksy · Kelly Holley · Gregory Sahagian · Tara Quesnell · Marc H. Feinberg · Gabrielle DeMaria · Mithila Vullaganti · Carolina Barnett-Tapia · Meg Mendoza · Mona Irannejad · Zaeem Siddiqi · Ali A. Habib · Thomas Ragole · Mamatha Pasnoor · Rebecca E. Traub · Manisha Chopra · Natalie S. Grover · Yara A. Park · Matthew Ka
9 January 2026 at 08:00

Nature Medicine, Published online: 09 January 2026; doi:10.1038/s41591-025-04170-z

Analysis of a placebo-controlled trial of a BCMA-targeting CAR-T cell therapy in patients with myasthenia gravis shows that CAR-T cell infusion selectively remodels the systemic immune environment, with elimination of BCMA-high plasma cells and activated plasmacytoid dendritic cells and changes in the autoreactive B cell repertoire.

BCMA-directed mRNA CAR T cell therapy for myasthenia gravis: a randomized, double-blind, placebo-controlled phase 2b trial

✇Nature Medicine
By: Tuan Vu · Hacer Durmus · Michael Rivner · Sheetal Shroff · Thomas Ragole · Bennett Myers · Mamatha Pasnoor · George Small · Chafic Karam · Mithila Vullaganti · Amanda Peltier · Gregory Sahagian · Marc H. Feinberg · Adam Slanksy · Carolina Barnett-Tapia · Zaeem Siddiqi · Kelly Gwathmey · Michael A. Badruddoja · Hafsa Kamboh · Rachel N. Ruggerie · Renee R. Fedak · C. Andrew Stewart · Metin Kurtoglu · Murat Kalayoglu · Michael Singer · Christopher M. Jewell · Milos D. Miljkovic · Mazen Dimachkie · Tahseen Mozaffar · James F. Howard Jr
9 January 2026 at 08:00

Nature Medicine, Published online: 09 January 2026; doi:10.1038/s41591-025-04171-y

In a randomized, double-blind, placebo-controlled trial comparing autologous mRNA-engineered BCMA-targeting CAR T cell therapy versus placebo in patients with generalized myasthenia gravis, a significantly higher percentage of patients exhibited a reduction in disease activity in the treatment arm than in the placebo arm.
Received — 9 January 2026 ⏭ Favourites

Sci-Reasoning: A Dataset Decoding AI Innovation Patterns

✇cs.AI, q-bio.NC updates on arXiv.org
By: Jiachen Liu · Maestro Harmon · Zechen Zhang
9 January 2026 at 13:00
arXiv:2601.04577v1 Announce Type: new Abstract: While AI innovation accelerates rapidly, the intellectual process behind breakthroughs -- how researchers identify gaps, synthesize prior work, and generate insights -- remains poorly understood. The lack of structured data on scientific reasoning hinders systematic analysis and development of AI research agents. We introduce Sci-Reasoning, the first dataset capturing the intellectual synthesis behind high-quality AI research. Using community-validated quality signals and an LLM-accelerated, human-verified pipeline, we trace Oral and Spotlight papers across NeurIPS, ICML, and ICLR (2023-2025) to its key predecessors, articulating specific reasoning links in a structured format. Our analysis identifies 15 distinct thinking patterns, with three dominant strategies accounting for 52.7%: Gap-Driven Reframing (24.2%), Cross-Domain Synthesis (18.0%), and Representation Shift (10.5%). The most powerful innovation recipes combine multiple patterns: Gap-Driven Reframing + Representation Shift, Cross-Domain Synthesis + Representation Shift, and Gap-Driven Reframing + Cross-Domain Synthesis. This dataset enables quantitative studies of scientific progress and provides structured reasoning trajectories for training the next generation AI research agents.

Self-MedRAG: a Self-Reflective Hybrid Retrieval-Augmented Generation Framework for Reliable Medical Question Answering

✇cs.AI, q-bio.NC updates on arXiv.org
By: Jessica Ryan · Alexander I. Gumilang · Robert Wiliam · Derwin Suhartono
9 January 2026 at 13:00
arXiv:2601.04531v1 Announce Type: cross Abstract: Large Language Models (LLMs) have demonstrated significant potential in medical Question Answering (QA), yet they remain prone to hallucinations and ungrounded reasoning, limiting their reliability in high-stakes clinical scenarios. While Retrieval-Augmented Generation (RAG) mitigates these issues by incorporating external knowledge, conventional single-shot retrieval often fails to resolve complex biomedical queries requiring multi-step inference. To address this, we propose Self-MedRAG, a self-reflective hybrid framework designed to mimic the iterative hypothesis-verification process of clinical reasoning. Self-MedRAG integrates a hybrid retrieval strategy, combining sparse (BM25) and dense (Contriever) retrievers via Reciprocal Rank Fusion (RRF) to maximize evidence coverage. It employs a generator to produce answers with supporting rationales, which are then assessed by a lightweight self-reflection module using Natural Language Inference (NLI) or LLM-based verification. If the rationale lacks sufficient evidentiary support, the system autonomously reformulates the query and iterates to refine the context. We evaluated Self-MedRAG on the MedQA and PubMedQA benchmarks. The results demonstrate that our hybrid retrieval approach significantly outperforms single-retriever baselines. Furthermore, the inclusion of the self-reflective loop yielded substantial gains, increasing accuracy on MedQA from 80.00% to 83.33% and on PubMedQA from 69.10% to 79.82%. These findings confirm that integrating hybrid retrieval with iterative, evidence-based self-reflection effectively reduces unsupported claims and enhances the clinical reliability of LLM-based systems.

Atlas 2 -- Foundation models for clinical deployment

✇cs.AI, q-bio.NC updates on arXiv.org
By: Maximilian Alber, Timo Milbich, Alexandra Carpen-Amarie, Stephan Tietz, Jonas Dippel, Lukas Muttenthaler, Beatriz Perez Cancer, Alessandro Benetti, Panos Korfiatis, Elias Eulig, J\'er\^ome L\"uscher, Jiasen Wu, Sayed Abid Hashimi, Gabriel Dernbach, Simon Schallenberg, Neelay Shah, Moritz Kr\"ugener, Aniruddh Jammoria, Jake Matras, Patrick Duffy, Matt Redlon, Philipp Jurmeister, David Horst, Lukas Ruff, Klaus-Robert M\"uller, Frederick Klauschen, Andrew Norgan
9 January 2026 at 13:00
arXiv:2601.05148v1 Announce Type: cross Abstract: Pathology foundation models substantially advanced the possibilities in computational pathology -- yet tradeoffs in terms of performance, robustness, and computational requirements remained, which limited their clinical deployment. In this report, we present Atlas 2, Atlas 2-B, and Atlas 2-S, three pathology vision foundation models which bridge these shortcomings by showing state-of-the-art performance in prediction performance, robustness, and resource efficiency in a comprehensive evaluation across eighty public benchmarks. Our models were trained on the largest pathology foundation model dataset to date comprising 5.5 million histopathology whole slide images, collected from three medical institutions Charit\'e - Universt\"atsmedizin Berlin, LMU Munich, and Mayo Clinic.

Multi-Modal AI for Remote Patient Monitoring in Cancer Care

✇cs.AI, q-bio.NC updates on arXiv.org
By: Yansong Liu · Ronnie Stafford · Pramit Khetrapal · Huriye Kocadag · Gra\c{c}a Carvalho · Patricia de Winter · Maryam Imran · Amelia Snook · Adamos Hadjivasiliou · D. Vijay Anand · Weining Lin · John Kelly · Yukun Zhou · Ivana Drobnjak
9 January 2026 at 13:00
arXiv:2512.00949v2 Announce Type: replace-cross Abstract: For patients undergoing systemic cancer therapy, the time between clinic visits is full of uncertainties and risks of unmonitored side effects. To bridge this gap in care, we developed and prospectively trialed a multi-modal AI framework for remote patient monitoring (RPM). This system integrates multi-modal data from the HALO-X platform, such as demographics, wearable sensors, daily surveys, and clinical events. Our observational trial is one of the largest of its kind and has collected over 2.1 million data points (6,080 patient-days) of monitoring from 84 patients. We developed and adapted a multi-modal AI model to handle the asynchronous and incomplete nature of real-world RPM data, forecasting a continuous risk of future adverse events. The model achieved an accuracy of 83.9% (AUROC=0.70). Notably, the model identified previous treatments, wellness check-ins, and daily maximum heart rate as key predictive features. A case study demonstrated the model's ability to provide early warnings by outputting escalating risk profiles prior to the event. This work establishes the feasibility of multi-modal AI RPM for cancer care and offers a path toward more proactive patient support.(Accepted at Europe NeurIPS 2025 Multimodal Representation Learning for Healthcare Workshop. Best Paper Poster Award.)

Developing an AI-Assisted Tool That Identifies Patients With Multimorbidity and Complex Polypharmacy to Improve the Process of Medication Reviews: Qualitative Interview and Focus Group Study

✇Journal of Medical Internet Research
By: Aseel S Abuzour · Samantha A Wilson · Alan A Woodall · Frances S Mair · Asra Aslam · Andrew Clegg · Eduard Shantsila · Mark Gabbay · Michael Abaho · Danushka Bollegala · Harriet Cant · Alan Griffiths · Layik Hama · Gary Leeming · Emma Lo · Simon Maskell · Maurice O'Connell · Olusegun Popoola · Sam Relton · Roy A Ruddle · Pieta Schofield · Matthew Sperrin · Tjeerd Van Staa · Iain Buchan · Lauren E Walker
9 January 2026 at 05:30
Background: Structured medication reviews (SMRs) are an essential component of medication optimization, especially for patients with multimorbidity and polypharmacy. However, the process remains challenging due to the complexities of patient data, time constraints, and the need for coordination among health care professionals (HCPs). This study explores HCPs’ perspectives on the integration of artificial intelligence (AI)–assisted tools to enhance the SMR process, with a focus on the potential benefits of and barriers to adoption. Objective: This study aims to identify the key user requirements for AI-assisted tools to improve the efficiency and effectiveness of SMRs, specifically for patients with multimorbidity, complex polypharmacy, and frailty. Methods: A qualitative study was conducted involving focus groups and semistructured interviews with HCPs and patients in the United Kingdom. Participants included physicians, pharmacists, clinical pharmacologists, psychiatrists from primary and secondary care, a policy maker, and patients with multimorbidity. Data were analyzed using a hybrid inductive and deductive thematic analysis approach to identify themes related to AI-assisted tool functionality, workflow integration, user-interface visualization, and usability in the SMR process. Results: Four major themes emerged from the analysis: innovative AI potential, optimizing electronic patient record visualization, functionality of the AI tool for SMRs, and facilitators of and barriers to AI tool implementation. HCPs identified the potential of AI to support patient identification and prioritizing those at risk of medication-related harm. AI-assisted tools were viewed as essential in detecting prescribing gaps, drug interactions, and patient risk trajectories over time. Participants emphasized the importance of presenting patient data in an intuitive format, with a patient interface for shared decision-making. Suggestions included color-coding blood results, highlighting critical medication reviews, and providing timelines of patient medical histories. HCPs stressed the need for AI tools to integrate seamlessly with existing electronic patient record systems and provide actionable insights without overwhelming users with excessive notifications or “pop-up” alerts. Factors influencing the uptake of AI-assisted tools included the need for user-friendly design, evidence of tool effectiveness (though some were skeptical about the predictive accuracy of AI models), and addressing concerns around digital exclusion. Conclusions: The findings highlight the potential for AI-assisted tools to streamline and optimize the SMR process, particularly for patients with multimorbidity and complex polypharmacy. However, successful implementation depends on addressing concerns related to workflow integration, user acceptance, and evidence of effectiveness. User-centered design is crucial to ensure that AI-assisted tools support HCPs in delivering high-quality, patient-centered care while minimizing cognitive overload and alert fatigue.

Intervention in Health Misinformation Using Large Language Models for Automated Detection, Thematic Analysis, and Inoculation: Case Study on COVID-19

✇Journal of Medical Internet Research
By: Samira Malek · Christopher Griffin · Robert D Fraleigh · Robert Lennon · Vishal Monga · Lijiang Shen
9 January 2026 at 02:15
Background: The rapid growth of social media as an information channel has enabled the swift spread of inaccurate or false health information, significantly impacting public health. This widespread dissemination of misinformation has caused confusion, eroded trust in health authorities, led to noncompliance with health guidelines, and encouraged risky health behaviors. Understanding the dynamics of misinformation on social media is essential for devising effective public health communication strategies. Objective: This study aims to present a comprehensive and automated approach that leverages Large Language Models (LLMs) and Machine Learning (ML) techniques to detect misinformation on social media, uncover the underlying causes and themes, and generate refutation arguments, facilitating control of its spread and promoting public health outcomes by inoculating people against health misinformation. Methods: We use two datasets to train three LLMs, namely BERT, T5, and GPT-2, to classify documents into two categories: misinformation and non-misinformation. Additionally, we employ a separate dataset to identify misinformation topics. To analyze these topics, we applied three topic modeling algorithms—Latent Dirichlet Allocation (LDA), Top2Vec, and BERTopic—and selected the optimal model based on performance evaluated across three metrics. Using a prompting approach, we extract sentence-level representations for the topics to uncover their underlying themes. Finally, we design a prompt text capable of identifying misinformation themes effectively. Codes are available at https://github.com/SamiraMalek/MDIP-MDIS. Results: The trained BERT model demonstrated exceptional performance, achieving 98% accuracy in classifying misinformation and non-misinformation, with a 44% reduction in false positive rates for AI-generated misinformation. Among the three topic modeling approaches employed, BERTopic outperformed the others, achieving the highest metrics with a Coherence Value (CV) of 0.41, Normalized Pointwise Mutual Information (NPMI) of -0.086, and Inverted RBO (IRBO) of 0.99. To address the issue of unclassified documents, we developed an algorithm to assign each document to its closest topic. Additionally, we proposed a novel method using prompt engineering to generate sentence-level representations for each topic, achieving a 99.6% approval rate as "appropriate" or "somewhat appropriate" by three independent raters. We further designed a prompt text to identify themes of misinformation topics and developed another prompt capable of detecting misinformation themes with 82% accuracy. Conclusions: This study presents a comprehensive and automated approach to addressing health misinformation on social media using advanced machine learning and natural language processing techniques. By leveraging large language models (LLMs) and prompt engineering, the system effectively detects misinformation, identifies underlying themes, and provides explanatory responses to combat its spread. The proposed method was tested on an English-language COVID-19–related dataset and has not been evaluated on real-world online social media data; the experiments were conducted offline.
Received — 8 January 2026 ⏭ Favourites

Organoids in translation: a bench-to-bedside framework for pancreatic cancer precision medicine

✇(Multiomics OR Omics) AND (Pancreatic)
By: Danial A Malik · Emily A S Schmieder · Gina Genova · Jeremey Gaskins · Robert C G Martin Ii
6 January 2026 at 19:00

J Transl Med. 2026 Jan 6. doi: 10.1186/s12967-025-07596-8. Online ahead of print.

ABSTRACT

INTRODUCTION: Pancreatic ductal adenocarcinoma (PDAC) is one of the most lethal malignancies with a 5-year survival rate of < 13%. Standard treatments such as FOLFIRINOX or gemcitabine/nab-paclitaxel yield modest response rates, underscoring the urgent need for precision oncology approaches. Patient-derived organoids (PDOs) preserve the genomic, phenotypic, and histopathological features of the source tumor and offer a promising platform for drug screening, biomarker development, and personalized therapy. However, a systematic evaluation of their translational capacities is lacking.

METHODS: A systematic review was conducted according to the PRISMA 2020 guidelines (PROSPERO registration pending) using PubMed, EMBASE, and Cochrane CENTRAL (December 10, 2024) to identify English-language PDAC PDO studies that incorporated therapeutic testing. Ninety-five studies met the inclusion criteria. Data extraction captured >75 variables per study, including spanning culture methodology, therapeutic profiling, biomarker integration, and clinical correlation. A 13-domain weighted Translatability Scoring Framework adapted from Wehling et al. assessed predictive validity, biomarker strength, pharmacogenetics, and clinical trial alignment. Scores ranged from 0 to 5 and were categorized as good (>4.0), moderate (3.0-4.0), or low (<3.0) translational potential.

RESULTS: Of the 95 studies, 70.5% have been published since 2021, reflecting the rapid growth in this field. The mean PDO generation success rate was 89.7%, with the primary tumor tissue being the predominant source (48.4%). Only 24.8% were directly linked to clinical trials and 5.3% incorporated multi-omic profiling. The median translatability score was 3.13 (range, 1.72-4.59): 45.3% of the studies had low translatability, 50.5% moderate, and only 4.2% had good translational potential. High-scoring studies consistently combine multi-omic biomarker platforms, in vivo validation, clinical outcome correlation, and prospective trial integration. Conversely, the weakest domains were pharmacogenetics, endpoint strategies, and biomarker validation, limiting their overall clinical relevance.

CONCLUSIONS: PDOs have demonstrated strong feasibility and in vitro clinical correlation in PDAC; however, their clinical translation remains constrained by limited multi-omic integration, absence of pharmacogenomic modeling, and sparse clinical trial embedding. Standardization of protocols, adoption of harmonized and clinically relevant endpoints, and systematic incorporation of biomarker-driven co-clinical trial frameworks are urgently needed to transition PDOs from promising experimental surrogates to validating precision oncology tools capable of informing therapeutic decision-making in PDAC.

PMID:41495743 | DOI:10.1186/s12967-025-07596-8

Received — 7 January 2026 ⏭ Favourites

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.

Yuan3.0 Flash: An Open Multimodal Large Language Model for Enterprise Applications

✇cs.AI, q-bio.NC updates on arXiv.org
By: YuanLab. ai · : · Shawn Wu · Sean Wang · Louie Li · Darcy Chen · Allen Wang · Jiangang Luo · Xudong Zhao · Joseph Shen · Gawain Ma · Jasper Jia · Marcus Mao · Claire Wang · Hunter He · Carol Wang · Zera Zhang · Jason Wang · Chonly Shen · Leo Zhang · Logan Chen · Qasim Meng · James Gong · Danied Zhao · Penn Zheng · Owen Zhu · Tong Yu
6 January 2026 at 13:00
arXiv:2601.01718v1 Announce Type: new Abstract: We introduce Yuan3.0 Flash, an open-source Mixture-of-Experts (MoE) MultiModal Large Language Model featuring 3.7B activated parameters and 40B total parameters, specifically designed to enhance performance on enterprise-oriented tasks while maintaining competitive capabilities on general-purpose tasks. To address the overthinking phenomenon commonly observed in Large Reasoning Models (LRMs), we propose Reflection-aware Adaptive Policy Optimization (RAPO), a novel RL training algorithm that effectively regulates overthinking behaviors. In enterprise-oriented tasks such as retrieval-augmented generation (RAG), complex table understanding, and summarization, Yuan3.0 Flash consistently achieves superior performance. Moreover, it also demonstrates strong reasoning capabilities in domains such as mathematics, science, etc., attaining accuracy comparable to frontier model while requiring only approximately 1/4 to 1/2 of the average tokens. Yuan3.0 Flash has been fully open-sourced to facilitate further research and real-world deployment: https://github.com/Yuan-lab-LLM/Yuan3.0.

XAI-MeD: Explainable Knowledge Guided Neuro-Symbolic Framework for Domain Generalization and Rare Class Detection in Medical Imaging

✇cs.AI, q-bio.NC updates on arXiv.org
By: Midhat Urooj · Ayan Banerjee · Sandeep Gupta
6 January 2026 at 13:00
arXiv:2601.02008v1 Announce Type: new Abstract: Explainability domain generalization and rare class reliability are critical challenges in medical AI where deep models often fail under real world distribution shifts and exhibit bias against infrequent clinical conditions This paper introduces XAIMeD an explainable medical AI framework that integrates clinically accurate expert knowledge into deep learning through a unified neuro symbolic architecture XAIMeD is designed to improve robustness under distribution shift enhance rare class sensitivity and deliver transparent clinically aligned interpretations The framework encodes clinical expertise as logical connectives over atomic medical propositions transforming them into machine checkable class specific rules Their diagnostic utility is quantified through weighted feature satisfaction scores enabling a symbolic reasoning branch that complements neural predictions A confidence weighted fusion integrates symbolic and deep outputs while a Hunt inspired adaptive routing mechanism guided by Entropy Imbalance Gain EIG and Rare Class Gini mitigates class imbalance high intra class variability and uncertainty We evaluate XAIMeD across diverse modalities on four challenging tasks i Seizure Onset Zone SOZ localization from rs fMRI ii Diabetic Retinopathy grading across 6 multicenter datasets demonstrate substantial performance improvements including 6 percent gains in cross domain generalization and a 10 percent improved rare class F1 score far outperforming state of the art deep learning baselines Ablation studies confirm that the clinically grounded symbolic components act as effective regularizers ensuring robustness to distribution shifts XAIMeD thus provides a principled clinically faithful and interpretable approach to multimodal medical AI.

MACA: A Framework for Distilling Trustworthy LLMs into Efficient Retrievers

✇cs.AI, q-bio.NC updates on arXiv.org
By: Satya Swaroop Gudipudi · Sahil Girhepuje · Ponnurangam Kumaraguru · Kristine Ma
6 January 2026 at 13:00
arXiv:2601.00926v1 Announce Type: cross Abstract: Modern enterprise retrieval systems must handle short, underspecified queries such as ``foreign transaction fee refund'' and ``recent check status''. In these cases, semantic nuance and metadata matter but per-query large language model (LLM) re-ranking and manual labeling are costly. We present Metadata-Aware Cross-Model Alignment (MACA), which distills a calibrated metadata aware LLM re-ranker into a compact student retriever, avoiding online LLM calls. A metadata-aware prompt verifies the teacher's trustworthiness by checking consistency under permutations and robustness to paraphrases, then supplies listwise scores, hard negatives, and calibrated relevance margins. The student trains with MACA's MetaFusion objective, which combines a metadata conditioned ranking loss with a cross model margin loss so it learns to push the correct answer above semantically similar candidates with mismatched topic, sub-topic, or entity. On a proprietary consumer banking FAQ corpus and BankFAQs, the MACA teacher surpasses a MAFA baseline at Accuracy@1 by five points on the proprietary set and three points on BankFAQs. MACA students substantially outperform pretrained encoders; e.g., on the proprietary corpus MiniLM Accuracy@1 improves from 0.23 to 0.48, while keeping inference free of LLM calls and supporting retrieval-augmented generation.

Digital Twin-Driven Communication-Efficient Federated Anomaly Detection for Industrial IoT

✇cs.AI, q-bio.NC updates on arXiv.org
By: Mohammed Ayalew Belay · Adil Rasheed · Pierluigi Salvo Rossi
6 January 2026 at 13:00
arXiv:2601.01701v1 Announce Type: cross Abstract: Anomaly detection is increasingly becoming crucial for maintaining the safety, reliability, and efficiency of industrial systems. Recently, with the advent of digital twins and data-driven decision-making, several statistical and machine-learning methods have been proposed. However, these methods face several challenges, such as dependence on only real sensor datasets, limited labeled data, high false alarm rates, and privacy concerns. To address these problems, we propose a suite of digital twin-integrated federated learning (DTFL) methods that enhance global model performance while preserving data privacy and communication efficiency. Specifically, we present five novel approaches: Digital Twin-Based Meta-Learning (DTML), Federated Parameter Fusion (FPF), Layer-wise Parameter Exchange (LPE), Cyclic Weight Adaptation (CWA), and Digital Twin Knowledge Distillation (DTKD). Each method introduces a unique mechanism to combine synthetic and real-world knowledge, balancing generalization with communication overhead. We conduct an extensive experiment using a publicly available cyber-physical anomaly detection dataset. For a target accuracy of 80%, CWA reaches the target in 33 rounds, FPF in 41 rounds, LPE in 48 rounds, and DTML in 87 rounds, whereas the standard FedAvg baseline and DTKD do not reach the target within 100 rounds. These results highlight substantial communication-efficiency gains (up to 62% fewer rounds than DTML and 31% fewer than LPE) and demonstrate that integrating DT knowledge into FL accelerates convergence to operationally meaningful accuracy thresholds for IIoT anomaly detection.
❌