Yinghao Zhu

Junyi Gao, Yinghao Zhu, Wenqing Wang et al.

The COVID-19 pandemic has posed a heavy burden to the healthcare system worldwide and caused huge social disruption and economic loss. Many deep learning models have been proposed to conduct clinical predictive tasks such as mortality prediction for COVID-19 patients in intensive care units using Electronic Health Record (EHR) data. Despite their initial success in certain clinical applications, there is currently a lack of benchmarking results to achieve a fair comparison so that we can select the optimal model for clinical use. Furthermore, there is a discrepancy between the formulation of traditional prediction tasks and real-world clinical practice in intensive care. To fill these gaps, we propose two clinical prediction tasks, Outcome-specific length-of-stay prediction and Early mortality prediction for COVID-19 patients in intensive care units. The two tasks are adapted from the naive length-of-stay and mortality prediction tasks to accommodate the clinical practice for COVID-19 patients. We propose fair, detailed, open-source data-preprocessing pipelines and evaluate 17 state-of-the-art predictive models on two tasks, including 5 machine learning models, 6 basic deep learning models and 6 deep learning predictive models specifically designed for EHR data. We provide benchmarking results using data from two real-world COVID-19 EHR datasets. One dataset is publicly available without needing any inquiry and another dataset can be accessed on request. We provide fair, reproducible benchmarking results for two tasks. We deploy all experiment results and models on an online platform. We also allow clinicians and researchers to upload their data to the platform and get quick prediction results using our trained models. We hope our efforts can further facilitate deep learning and machine learning research for COVID-19 predictive modeling.

Yinghao Zhu, Zixiang Wang, Long He et al.

Electronic Health Records (EHRs) contain a wealth of patient data; however, the sparsity of EHRs data often presents significant challenges for predictive modeling. Conventional imputation methods inadequately distinguish between real and imputed data, leading to potential inaccuracies of patient representations. To address these issues, we introduce PRISM, a framework that indirectly imputes data by leveraging prototype representations of similar patients, thus ensuring compact representations that preserve patient information. PRISM also includes a feature confidence learner module, which evaluates the reliability of each feature considering missing statuses. Additionally, PRISM introduces a new patient similarity metric that accounts for feature confidence, avoiding over-reliance on imprecise imputed values. Our extensive experiments on the MIMIC-III, MIMIC-IV, PhysioNet Challenge 2012, eICU datasets demonstrate PRISM's superior performance in predicting in-hospital mortality and 30-day readmission tasks, showcasing its effectiveness in handling EHR data sparsity. For the sake of reproducibility and further research, we have publicly released the code at https://github.com/yhzhu99/PRISM.

Zhongying Deng, Cheng Tang, Ziyan Huang et al. · pku

Foundation models have demonstrated remarkable success across diverse domains and tasks, primarily due to the thrive of large-scale, diverse, and high-quality datasets. However, in the field of medical imaging, the curation and assembling of such medical datasets are highly challenging due to the reliance on clinical expertise and strict ethical and privacy constraints, resulting in a scarcity of large-scale unified medical datasets and hindering the development of powerful medical foundation models. In this work, we present the largest survey to date of medical image datasets, covering over 1,000 open-access datasets with a systematic catalog of their modalities, tasks, anatomies, annotations, limitations, and potential for integration. Our analysis exposes a landscape that is modest in scale, fragmented across narrowly scoped tasks, and unevenly distributed across organs and modalities, which in turn limits the utility of existing medical image datasets for developing versatile and robust medical foundation models. To turn fragmentation into scale, we propose a metadata-driven fusion paradigm (MDFP) that integrates public datasets with shared modalities or tasks, thereby transforming multiple small data silos into larger, more coherent resources. Building on MDFP, we release an interactive discovery portal that enables end-to-end, automated medical image dataset integration, and compile all surveyed datasets into a unified, structured table that clearly summarizes their key characteristics and provides reference links, offering the community an accessible and comprehensive repository. By charting the current terrain and offering a principled path to dataset consolidation, our survey provides a practical roadmap for scaling medical imaging corpora, supporting faster data discovery, more principled dataset creation, and more capable medical foundation models.

Yinghao Zhu, Junyi Gao, Zixiang Wang et al.

Large Language Models (LLMs) are increasingly deployed in medicine. However, their utility in non-generative clinical prediction, often presumed inferior to specialized models, remains under-evaluated, leading to ongoing debate within the field and potential for misuse, misunderstanding, or over-reliance due to a lack of systematic benchmarking. Our ClinicRealm study addresses this by benchmarking 15 GPT-style LLMs, 5 BERT-style models, and 11 traditional methods on unstructured clinical notes and structured Electronic Health Records (EHR), while also assessing their reasoning, reliability, and fairness. Key findings reveal a significant shift: for clinical note predictions, leading LLMs (e.g., DeepSeek-V3.1-Think, GPT-5) in zero-shot settings now decisively outperform finetuned BERT models. On structured EHRs, while specialized models excel with ample data, advanced LLMs (e.g., GPT-5, DeepSeek-V3.1-Think) show potent zero-shot capabilities, often surpassing conventional models in data-scarce settings. Notably, leading open-source LLMs can match or exceed proprietary counterparts. These results provide compelling evidence that modern LLMs are competitive tools for non-generative clinical prediction, particularly with unstructured text and offering data-efficient structured data options, thus necessitating a re-evaluation of model selection strategies. This research should serve as an important insight for medical informaticists, AI developers, and clinical researchers, potentially prompting a reassessment of current assumptions and inspiring new approaches to LLM application in predictive healthcare.

Zhongji Zhang, Yuhang Wang, Yinghao Zhu et al.

Emerging diseases present challenges in symptom recognition and timely clinical intervention due to limited available information. An effective prognostic model could assist physicians in making accurate diagnoses and designing personalized treatment plans to prevent adverse outcomes. However, in the early stages of disease emergence, several factors hamper model development: limited data collection, insufficient clinical experience, and privacy and ethical concerns restrict data availability and complicate accurate label assignment. Furthermore, Electronic Medical Record (EMR) data from different diseases or sources often exhibit significant cross-dataset feature misalignment, severely impacting the effectiveness of deep learning models. We present a domain-invariant representation learning method that constructs a transition model between source and target datasets. By constraining the distribution shift of features generated across different domains, we capture domain-invariant features specifically relevant to downstream tasks, developing a unified domain-invariant encoder that achieves better feature representation across various task domains. Experimental results across multiple target tasks demonstrate that our proposed model surpasses competing baseline methods and achieves faster training convergence, particularly when working with limited data. Extensive experiments validate our method's effectiveness in providing more accurate predictions for emerging pandemics and other diseases. Code is publicly available at https://github.com/wang1yuhang/domain_invariant_network.

Chaohe Zhang, Xu Chu, Liantao Ma et al.

Multimodal electronic health record (EHR) data are widely used in clinical applications. Conventional methods usually assume that each sample (patient) is associated with the unified observed modalities, and all modalities are available for each sample. However, missing modality caused by various clinical and social reasons is a common issue in real-world clinical scenarios. Existing methods mostly rely on solving a generative model that learns a mapping from the latent space to the original input space, which is an unstable ill-posed inverse problem. To relieve the underdetermined system, we propose a model solving a direct problem, dubbed learning with Missing Modalities in Multimodal healthcare data (M3Care). M3Care is an end-to-end model compensating the missing information of the patients with missing modalities to perform clinical analysis. Instead of generating raw missing data, M3Care imputes the task-related information of the missing modalities in the latent space by the auxiliary information from each patient's similar neighbors, measured by a task-guided modality-adaptive similarity metric, and thence conducts the clinical tasks. The task-guided modality-adaptive similarity metric utilizes the uncensored modalities of the patient and the other patients who also have the same uncensored modalities to find similar patients. Experiments on real-world datasets show that M3Care outperforms the state-of-the-art baselines. Moreover, the findings discovered by M3Care are consistent with experts and medical knowledge, demonstrating the capability and the potential of providing useful insights and explanations.

Yinghao Zhu, Jingkun An, Enshen Zhou et al. · tencent-ai

In the data-driven artificial intelligence paradigm, models heavily rely on large amounts of training data. However, factors like sampling distribution imbalance can lead to issues of bias and unfairness in healthcare data. Sensitive attributes, such as race, gender, age, and medical condition, are characteristics of individuals that are commonly associated with discrimination or bias. In healthcare AI, these attributes can play a significant role in determining the quality of care that individuals receive. For example, minority groups often receive fewer procedures and poorer-quality medical care than white individuals in US. Therefore, detecting and mitigating bias in data is crucial to enhancing health equity. Bias mitigation methods include pre-processing, in-processing, and post-processing. Among them, Reweighting (RW) is a widely used pre-processing method that performs well in balancing machine learning performance and fairness performance. RW adjusts the weights for samples within each (group, label) combination, where these weights are utilized in loss functions. However, RW is limited to considering only a single sensitive attribute when mitigating bias and assumes that each sensitive attribute is equally important. This may result in potential inaccuracies when addressing intersectional bias. To address these limitations, we propose M3Fair, a multi-level and multi-sensitive-attribute reweighting method by extending the RW method to multiple sensitive attributes at multiple levels. Our experiments on real-world datasets show that the approach is effective, straightforward, and generalizable in addressing the healthcare fairness issues.

Weibin Liao, Yinghao Zhu, Xinyuan Wang et al.

UNet and its variants have been widely used in medical image segmentation. However, these models, especially those based on Transformer architectures, pose challenges due to their large number of parameters and computational loads, making them unsuitable for mobile health applications. Recently, State Space Models (SSMs), exemplified by Mamba, have emerged as competitive alternatives to CNN and Transformer architectures. Building upon this, we employ Mamba as a lightweight substitute for CNN and Transformer within UNet, aiming at tackling challenges stemming from computational resource limitations in real medical settings. To this end, we introduce the Lightweight Mamba UNet (LightM-UNet) that integrates Mamba and UNet in a lightweight framework. Specifically, LightM-UNet leverages the Residual Vision Mamba Layer in a pure Mamba fashion to extract deep semantic features and model long-range spatial dependencies, with linear computational complexity. Extensive experiments conducted on two real-world 2D/3D datasets demonstrate that LightM-UNet surpasses existing state-of-the-art literature. Notably, when compared to the renowned nnU-Net, LightM-UNet achieves superior segmentation performance while drastically reducing parameter and computation costs by 116x and 21x, respectively. This highlights the potential of Mamba in facilitating model lightweighting. Our code implementation is publicly available at https://github.com/MrBlankness/LightM-UNet.

Ziyi He, Yushi Feng, Shuangyu Yang et al.

Dental triage is a safety-critical clinical routing task that requires integrating multimodal clinical information (e.g., patient complaints and radiographic evidence) to determine complete referral plans. We present Dental-TriageBench, the first expert-annotated benchmark for reasoning-driven multimodal dental triage. Built from authentic outpatient workflows, it contains 246 de-identified cases annotated with expert-authored golden reasoning trajectories, together with hierarchical triage labels. We benchmark 19 proprietary, open-source, and medical-domain MLLMs against three junior dentists serving as the human baseline, and find a substantial human--model gap, on fine-grained treatment-level triage. Further analyses show that accurate triage requires both complaint and OPG information, and that model errors concentrate on cases with multiple referral domains, where MLLMs tend to produce overly narrow referral sets and omission-heavy errors. Dental-TriageBench provides a realistic testbed for developing multimodal clinical AI systems that are more clinically grounded, coverage-aware, and safer for downstream care.

Yinghao Zhu, Ziyi He, Haoran Hu et al.

The rapid advancement of Large Language Models (LLMs) has stimulated interest in multi-agent collaboration for addressing complex medical tasks. However, the practical advantages of multi-agent collaboration approaches remain insufficiently understood. Existing evaluations often lack generalizability, failing to cover diverse tasks reflective of real-world clinical practice, and frequently omit rigorous comparisons against both single-LLM-based and established conventional methods. To address this critical gap, we introduce MedAgentBoard, a comprehensive benchmark for the systematic evaluation of multi-agent collaboration, single-LLM, and conventional approaches. MedAgentBoard encompasses four diverse medical task categories: (1) medical (visual) question answering, (2) lay summary generation, (3) structured Electronic Health Record (EHR) predictive modeling, and (4) clinical workflow automation, across text, medical images, and structured EHR data. Our extensive experiments reveal a nuanced landscape: while multi-agent collaboration demonstrates benefits in specific scenarios, such as enhancing task completeness in clinical workflow automation, it does not consistently outperform advanced single LLMs (e.g., in textual medical QA) or, critically, specialized conventional methods that generally maintain better performance in tasks like medical VQA and EHR-based prediction. MedAgentBoard offers a vital resource and actionable insights, emphasizing the necessity of a task-specific, evidence-based approach to selecting and developing AI solutions in medicine. It underscores that the inherent complexity and overhead of multi-agent collaboration must be carefully weighed against tangible performance gains. All code, datasets, detailed prompts, and experimental results are open-sourced at https://medagentboard.netlify.app/.

Jingkun An, Yinghao Zhu, Zongjian Li et al.

Text-to-Image (T2I) diffusion models have achieved remarkable success in image generation. Despite their progress, challenges remain in both prompt-following ability, image quality and lack of high-quality datasets, which are essential for refining these models. As acquiring labeled data is costly, we introduce AGFSync, a framework that enhances T2I diffusion models through Direct Preference Optimization (DPO) in a fully AI-driven approach. AGFSync utilizes Vision-Language Models (VLM) to assess image quality across style, coherence, and aesthetics, generating feedback data within an AI-driven loop. By applying AGFSync to leading T2I models such as SD v1.4, v1.5, and SDXL-base, our extensive experiments on the TIFA dataset demonstrate notable improvements in VQA scores, aesthetic evaluations, and performance on the HPSv2 benchmark, consistently outperforming the base models. AGFSync's method of refining T2I diffusion models paves the way for scalable alignment techniques. Our code and dataset are publicly available at https://anjingkun.github.io/AGFSync.

Simiao Liu, Li Zhang, Fang Liu et al.

Modern software ecosystems face a rapidly growing number of disclosed vulnerabilities, increasing the need for automated repair techniques that can operate reliably at repository scale. Although Large Language Model (LLM)-based agents have recently shown promise for automated vulnerability repair (AVR), most existing systems still treat repair as a single generation step over the currently visible code context. As a result, they lack a persistent mechanism for reusing prior fixes or learning from failed validation attempts, which limits their effectiveness on complex, multi-file repair tasks. We present MemRepair, a memory-augmented agentic framework that formulates vulnerability repair as an iterative, experience-driven process. MemRepair combines three complementary memory layers, i.e., History-Fix, Security-Pattern, and Refinement-Trajectory memories, with a dynamic feedback-driven refinement loop. This design allows the agent to retrieve repository-specific repair conventions, apply reusable security defenses, and exploit prior "failure-to-success" trajectories to revise semantically invalid patches based on runtime evidence. We evaluate MemRepair on three representative repository-level vulnerability repair benchmarks: SEC-Bench, PatchEval (Python, Go, JavaScript), and the C++ subset of Multi-SWE-bench. MemRepair achieves state-of-the-art resolution rates of 58.0%, 58.2%, and 30.58%, respectively, outperforming strong general-purpose agents such as OpenHands and SWE-agent, as well as the specialized AVR tool InfCode-C++, while maintaining competitive repair cost. These results show that persistent, hierarchical repair memory can substantially improve the reliability of agentic vulnerability repair across diverse languages and repository settings.

Simiao Liu, Fang Liu, Li Zhang et al.

Large language model (LLM) agents are increasingly used for automated vulnerability repair (AVR), where repository-level reasoning enables them to inspect context and produce source-code patches. However, recent empirical results show that these agents still struggle with real-world vulnerabilities. Their main failure mode is semantic misunderstanding: choosing a repair direction that does not match the root cause. We identify two reasons for this gap. Existing agents usually reason from the failing execution alone. A crash report can pinpoint where the program failed, but it does not reveal which variable or state transition, among many candidates near the fault site, separates the crashing behavior from safe execution. As a result, agents often produce symptom-oriented patches instead of causal fixes. Moreover, evidence collected for one vulnerability is rarely retained, so similar cases in later repositories must be diagnosed again from scratch. We present ContraFix, an agentic AVR framework that couples differential runtime evidence with reusable repair skills. Its Mutator constructs PoC variants that straddle the failure boundary; its Analyzer inserts state probes around the fault region and summarizes divergences between crashing and non-crashing executions into a repair specification; and its Patcher converts the specification into verified source patches. Each successful repair updates a two-track skill base containing repair specifications and mutation strategies, which are retrieved through a three-tier policy for future instances. On SEC-Bench (C/C++, 200 instances) and PatchEval (Go, Python, JavaScript, 225 instances), ContraFix with GPT-5-mini resolves 84.0% and 73.8% of the tasks, respectively, achieving state-of-the-art performance on both benchmarks while costing less than one-third of the strongest comparable baseline.

Fang Liu, Simiao Liu, Yinghao Zhu et al.

Identifying and addressing security issues during the early phase of the development lifecycle is critical for mitigating the long-term negative impacts on software systems. Code review serves as an effective practice that enables developers to check their teammates' code before integration into the codebase. To streamline the generation of review comments, various automated code review approaches have been proposed, where LLM-based methods have significantly advanced the capabilities of automated review generation. However, existing models primarily focus on general-purpose code review, their effectiveness in identifying and addressing security-related issues remains underexplored. Moreover, adapting existing code review approaches to target security issues faces substantial challenges, including data scarcity and inadequate evaluation metrics. To address these limitations, we propose SecureReviewer, a new approach designed for enhancing LLMs' ability to identify and resolve security-related issues during code review. Specifically, we first construct a dataset tailored for training and evaluating secure code review capabilities. Leveraging this dataset, we fine-tune LLMs to generate code review comments that can effectively identify security issues and provide fix suggestions with our proposed secure-aware fine-tuning strategy. To mitigate hallucination in LLMs and enhance the reliability of their outputs, we integrate the RAG technique, which grounds the generated comments in domain-specific security knowledge. Additionally, we introduce SecureBLEU, a new evaluation metric designed to assess the effectiveness of review comments in addressing security issues. Experimental results demonstrate that SecureReviewer outperforms state-of-the-art baselines in both security issue detection accuracy and the overall quality and practical utility of generated review comments.

Xichen Zhang, Ziyi He, Yinghao Zhu et al.

Search agents have emerged as a pivotal paradigm for solving open-ended, knowledge-intensive reasoning tasks. However, training these agents via Reinforcement Learning (RL) faces a critical dilemma: interacting with live commercial Web APIs is prohibitively expensive, while relying on static data snapshots often introduces noise due to data misalignment. This misalignment generates corrupted reward signals that destabilize training by penalizing correct reasoning or rewarding hallucination. To address this, we propose SearchGym, a simulation environment designed to bootstrap robust search agents. SearchGym employs a rigorous generative pipeline to construct a verifiable knowledge graph and an aligned document corpus, ensuring that every reasoning task is factually grounded and strictly solvable. Building on this controllable environment, we introduce SearchGym-RL, a curriculum learning methodology that progressively optimizes agent policies through purified feedback, evolving from basic interactions to complex, long-horizon planning. Extensive experiments across the Llama and Qwen families demonstrate strong Sim-to-Real generalization. Notably, our Qwen2.5-7B-Base model trained within SearchGym surpasses the web-enhanced ASearcher baseline across nine diverse benchmarks by an average relative margin of 10.6%. Our results validate that high-fidelity simulation serves as a scalable and highly cost-effective methodology for developing capable search agents.

Yue Fang, Yuxin Guo, Jiaran Gao et al.

Improving large language models (LLMs) for electronic health record (EHR) reasoning is essential for enabling accurate and generalizable clinical predictions. While LLMs excel at medical text understanding, they underperform on EHR-based prediction tasks due to challenges in modeling temporally structured, high-dimensional data. Existing approaches often rely on hybrid paradigms, where LLMs serve merely as frozen prior retrievers while downstream deep learning (DL) models handle prediction, failing to improve the LLM's intrinsic reasoning capacity and inheriting the generalization limitations of DL models. To this end, we propose EAG-RL, a novel two-stage training framework designed to intrinsically enhance LLMs' EHR reasoning ability through expert attention guidance, where expert EHR models refer to task-specific DL models trained on EHR data. Concretely, EAG-RL first constructs high-quality, stepwise reasoning trajectories using expert-guided Monte Carlo Tree Search to effectively initialize the LLM's policy. Then, EAG-RL further optimizes the policy via reinforcement learning by aligning the LLM's attention with clinically salient features identified by expert EHR models. Extensive experiments on two real-world EHR datasets show that EAG-RL improves the intrinsic EHR reasoning ability of LLMs by an average of 14.62%, while also enhancing robustness to feature perturbations and generalization to unseen clinical domains. These results demonstrate the practical potential of EAG-RL for real-world deployment in clinical prediction tasks. Our code have been available at https://github.com/devilran6/EAG-RL.

Weibin Liao, Tianlong Wang, Yinghao Zhu et al.

Medical Lay Language Generation (MLLG) plays a vital role in improving the accessibility of complex scientific content for broader audiences. Recent literature to MLLG commonly employ parameter-efficient fine-tuning methods such as Low-Rank Adaptation (LoRA) to fine-tuning large language models (LLMs) using paired expert-lay language datasets. However, LoRA struggles with the challenges posed by multi-source heterogeneous MLLG datasets. Specifically, through a series of exploratory experiments, we reveal that standard LoRA fail to meet the requirement for semantic fidelity and diverse lay-style generation in MLLG task. To address these limitations, we propose Magical, an asymmetric LoRA architecture tailored for MLLG under heterogeneous data scenarios. Magical employs a shared matrix $A$ for abstractive summarization, along with multiple isolated matrices $B$ for diverse lay-style generation. To preserve semantic fidelity during the lay language generation process, Magical introduces a Semantic Invariance Constraint to mitigate semantic subspace shifts on matrix $A$. Furthermore, to better adapt to diverse lay-style generation, Magical incorporates the Recommendation-guided Switch, an externally interface to prompt the LLM to switch between different matrices $B$. Experimental results on three real-world lay language generation datasets demonstrate that Magical consistently outperforms prompt-based methods, vanilla LoRA, and its recent variants, while also reducing trainable parameters by 31.66%. Our code is publicly available at https://github.com/tianlwang/Magical.git.

Xiaoyan Su, Yinghao Zhu, Run Li

In the past, research on a single low dimensional activation function in networks has led to internal covariate shift and gradient deviation problems. A relatively small research area is how to use function combinations to provide property completion for a single activation function application. We propose a network adversarial method to address the aforementioned challenges. This is the first method to use different activation functions in a network. Based on the existing activation functions in the current network, an adversarial function with opposite derivative image properties is constructed, and the two are alternately used as activation functions for different network layers. For complex situations, we propose a method of high-dimensional function graph decomposition(HD-FGD), which divides it into different parts and then passes through a linear layer. After integrating the inverse of the partial derivatives of each decomposed term, we obtain its adversarial function by referring to the computational rules of the decomposition process. The use of network adversarial methods or the use of HD-FGD alone can effectively replace the traditional MLP+activation function mode. Through the above methods, we have achieved a substantial improvement over standard activation functions regarding both training efficiency and predictive accuracy. The article addresses the adversarial issues associated with several prevalent activation functions, presenting alternatives that can be seamlessly integrated into existing models without any adverse effects. We will release the code as open source after the conference review process is completed.

Yinghao Zhu, Zixiang Wang, Junyi Gao et al.

The inherent complexity of structured longitudinal Electronic Health Records (EHR) data poses a significant challenge when integrated with Large Language Models (LLMs), which are traditionally tailored for natural language processing. Motivated by the urgent need for swift decision-making during new disease outbreaks, where traditional predictive models often fail due to a lack of historical data, this research investigates the adaptability of LLMs, like GPT-4, to EHR data. We particularly focus on their zero-shot capabilities, which enable them to make predictions in scenarios in which they haven't been explicitly trained. In response to the longitudinal, sparse, and knowledge-infused nature of EHR data, our prompting approach involves taking into account specific EHR characteristics such as units and reference ranges, and employing an in-context learning strategy that aligns with clinical contexts. Our comprehensive experiments on the MIMIC-IV and TJH datasets demonstrate that with our elaborately designed prompting framework, LLMs can improve prediction performance in key tasks such as mortality, length-of-stay, and 30-day readmission by about 35\%, surpassing ML models in few-shot settings. Our research underscores the potential of LLMs in enhancing clinical decision-making, especially in urgent healthcare situations like the outbreak of emerging diseases with no labeled data. The code is publicly available at https://github.com/yhzhu99/llm4healthcare for reproducibility.

Yinghao Zhu, Changyu Ren, Shiyun Xie et al.

The integration of multimodal Electronic Health Records (EHR) data has significantly improved clinical predictive capabilities. Leveraging clinical notes and multivariate time-series EHR, existing models often lack the medical context relevent to clinical tasks, prompting the incorporation of external knowledge, particularly from the knowledge graph (KG). Previous approaches with KG knowledge have primarily focused on structured knowledge extraction, neglecting unstructured data modalities and semantic high dimensional medical knowledge. In response, we propose REALM, a Retrieval-Augmented Generation (RAG) driven framework to enhance multimodal EHR representations that address these limitations. Firstly, we apply Large Language Model (LLM) to encode long context clinical notes and GRU model to encode time-series EHR data. Secondly, we prompt LLM to extract task-relevant medical entities and match entities in professionally labeled external knowledge graph (PrimeKG) with corresponding medical knowledge. By matching and aligning with clinical standards, our framework eliminates hallucinations and ensures consistency. Lastly, we propose an adaptive multimodal fusion network to integrate extracted knowledge with multimodal EHR data. Our extensive experiments on MIMIC-III mortality and readmission tasks showcase the superior performance of our REALM framework over baselines, emphasizing the effectiveness of each module. REALM framework contributes to refining the use of multimodal EHR data in healthcare and bridging the gap with nuanced medical context essential for informed clinical predictions.

Zirui Zhao, Junchao Xia, Si Wu et al.

In recent years, researchers have increasingly sought batteries as an efficient and cost-effective solution for energy storage and supply, owing to their high energy density, low cost, and environmental resilience. However, the issue of dendrite growth has emerged as a significant obstacle in battery development. Excessive dendrite growth during charging and discharging processes can lead to battery short-circuiting, degradation of electrochemical performance, reduced cycle life, and abnormal exothermic events. Consequently, understanding the dendrite growth process has become a key challenge for researchers. In this study, we investigated dendrite growth mechanisms in batteries using a combined machine learning approach, specifically a two-dimensional artificial convolutional neural network (CNN) model, along with computational methods. We developed two distinct computer models to predict dendrite growth in batteries. The CNN-1 model employs standard convolutional neural network techniques for dendritic growth prediction, while CNN-2 integrates additional physical parameters to enhance model robustness. Our results demonstrate that CNN-2 significantly enhances prediction accuracy, offering deeper insights into the impact of physical factors on dendritic growth. This improved model effectively captures the dynamic nature of dendrite formation, exhibiting high accuracy and sensitivity. These findings contribute to the advancement of safer and more reliable energy storage systems.

Yinghao Zhu, Yifan Qi, Zixiang Wang et al.

The rapid proliferation of scientific knowledge presents a grand challenge: transforming this vast repository of information into an active engine for discovery, especially in high-stakes domains like healthcare. Current AI agents, however, are constrained by static, predefined strategies, limiting their ability to navigate the complex, evolving ecosystem of scientific research. This paper introduces HealthFlow, a self-evolving AI agent that overcomes this limitation through a novel meta-level evolution mechanism. HealthFlow autonomously refines its high-level problem-solving policies by distilling procedural successes and failures into a durable, structured knowledge base, enabling it to learn not just how to use tools, but how to strategize. To anchor our research and provide a community resource, we introduce EHRFlowBench, a new benchmark featuring complex health data analysis tasks systematically derived from peer-reviewed scientific literature. Our experiments demonstrate that HealthFlow's self-evolving approach significantly outperforms state-of-the-art agent frameworks. This work offers a new paradigm for intelligent systems that can learn to operationalize the procedural knowledge embedded in scientific content, marking a critical step toward more autonomous and effective AI for healthcare scientific discovery.

Weibin Liao, Yinghao Zhu, Zhongji Zhang et al.

Analyzing the health status of patients based on Electronic Health Records (EHR) is a fundamental research problem in medical informatics. The presence of extensive missing values in EHR makes it challenging for deep neural networks (DNNs) to directly model the patient's health status. Existing DNNs training protocols, including Impute-then-Regress Procedure and Jointly Optimizing of Impute-n-Regress Procedure, require the additional imputation models to reconstruction missing values. However, Impute-then-Regress Procedure introduces the risk of injecting imputed, non-real data into downstream clinical prediction tasks, resulting in power loss, biased estimation, and poorly performing models, while Jointly Optimizing of Impute-n-Regress Procedure is also difficult to generalize due to the complex optimization space and demanding data requirements. Inspired by the recent advanced literature of learnable prompt in the fields of NLP and CV, in this work, we rethought the necessity of the imputation model in downstream clinical tasks, and proposed Learnable Prompt as Pseudo-Imputation (PAI) as a new training protocol to assist EHR analysis. PAI no longer introduces any imputed data but constructs a learnable prompt to model the implicit preferences of the downstream model for missing values, resulting in a significant performance improvement for all state-of-the-arts EHR analysis models on four real-world datasets across two clinical prediction tasks. Further experimental analysis indicates that PAI exhibits higher robustness in situations of data insufficiency and high missing rates. More importantly, as a plug-and-play protocol, PAI can be easily integrated into any existing or even imperceptible future EHR analysis models.

Simiao Liu, Fang Liu, Liehao Li et al.

Automated issue solving seeks to autonomously identify and repair defective code snippets across an entire codebase. SWE-Bench has emerged as the most widely adopted benchmark for evaluating progress in this area. While LLM-based agentic tools show great promise, they still fail on a substantial portion of tasks. Moreover, current evaluations primarily report aggregate issue-solving rates, which obscure the underlying causes of success and failure, making it challenging to diagnose model weaknesses or guide targeted improvements. To bridge this gap, we first analyze the performance and efficiency of three SOTA tools, spanning both pipeline-based and agentic architectures, in automated issue solving tasks of SWE-Bench-Verified under varying task characteristics. Furthermore, to move from high-level performance metrics to underlying cause analysis, we conducted a systematic manual analysis of 150 failed instances. From this analysis, we developed a comprehensive taxonomy of failure modes comprising 3 primary phases, 9 main categories, and 25 fine-grained subcategories. Then we systematically analyze the distribution of the identified failure modes, the results reveal distinct failure fingerprints between the two architectural paradigms, with the majority of agentic failures stemming from flawed reasoning and cognitive deadlocks. Motivated by these insights, we propose a collaborative Expert-Executor framework. It introduces a supervisory Expert agent tasked with providing strategic oversight and course-correction for a primary Executor agent. This architecture is designed to correct flawed reasoning and break the cognitive deadlocks that frequently lead to failure. Experiments show that our framework solves 22.2% of previously intractable issues for a leading single agent. These findings pave the way for building more robust agents through diagnostic evaluation and collaborative design.

Xichen Zhang, Sitong Wu, Yinghao Zhu et al.

Reinforcement learning from verifiable rewards has emerged as a powerful technique for enhancing the complex reasoning abilities of Large Language Models (LLMs). However, these methods are fundamentally constrained by the ''learning cliff'' phenomenon: when faced with problems far beyond their current capabilities, models consistently fail, yielding a persistent zero-reward signal. In policy optimization algorithms like GRPO, this collapses the advantage calculation to zero, rendering these difficult problems invisible to the learning gradient and stalling progress. To overcome this, we introduce Scaf-GRPO (Scaffolded Group Relative Policy Optimization), a progressive training framework that strategically provides minimal guidance only when a model's independent learning has plateaued. The framework first diagnoses learning stagnation and then intervenes by injecting tiered in-prompt hints, ranging from abstract concepts to concrete steps, enabling the model to construct a valid solution by itself. Extensive experiments on challenging mathematics benchmarks demonstrate Scaf-GRPO's effectiveness, boosting the pass@1 score of the Qwen2.5-Math-7B model on the AIME24 benchmark by a relative 44.3% over a vanilla GRPO baseline. This result demonstrates our framework provides a robust and effective methodology for unlocking a model's ability to solve problems previously beyond its reach, a critical step towards extending the frontier of autonomous reasoning in LLM.

Hongjun Liu, Yinghao Zhu, Yuhui Wang et al.

Recent progress in multimodal large language models (MLLMs) has demonstrated promising performance on medical benchmarks and in preliminary trials as clinical assistants. Yet, our pilot audit of diagnostic cases uncovers a critical failure mode: instability in early evidence interpretation precedes hallucination, creating branching reasoning trajectories that cascade into globally inconsistent conclusions. This highlights the need for clinical reasoning agents that constrain stochasticity and hallucination while producing auditable decision flows. We introduce MedMMV, a controllable multimodal multi-agent framework for reliable and verifiable clinical reasoning. MedMMV stabilizes reasoning through diversified short rollouts, grounds intermediate steps in a structured evidence graph under the supervision of a Hallucination Detector, and aggregates candidate paths with a Combined Uncertainty scorer. On six medical benchmarks, MedMMV improves accuracy by up to 12.7% and, more critically, demonstrates superior reliability. Blind physician evaluations confirm that MedMMV substantially increases reasoning truthfulness without sacrificing informational content. By controlling instability through a verifiable, multi-agent process, our framework provides a robust path toward deploying trustworthy AI systems in high-stakes domains like clinical decision support.

Xichen Zhang, Sitong Wu, Haoru Tan et al.

The long chain-of-thought (LongCoT) capability is central to the recent breakthroughs achieved by large language models in complex reasoning tasks. However, the accompanying issue of ''underthinking'', where models exhibit shallow reasoning by frequently switching thoughts without sufficient exploration, limits both performance and token efficiency. To address this problem, we propose a simple yet effective reasoning strategy: the SmartSwitch inference framework. This framework can be easily integrated into any large language model as a plug-and-play solution, continuously monitoring the model's reasoning process to detect underthinking and guide it toward deeper exploration of promising but overlooked thoughts. Specifically, the perception module identifies points where thoughts switch and evaluates the potential of the preceding thought using an off-the-shelf process reward model (PRM). If a high-potential thought is found to be prematurely abandoned, the intervention module interrupts the ongoing inference, backtracks to the point before the switch, and inserts a "deepening prompt" to encourage further exploration along that promising path. Extensive experiments on challenging mathematical reasoning benchmarks demonstrate that our method significantly enhances the performance of various large language models of different sizes.

Lei Gu, Yinghao Zhu, Haoran Sang et al.

While large language model (LLM)-based multi-agent systems show promise in simulating medical consultations, their evaluation is often confined to final-answer accuracy. This practice treats their internal collaborative processes as opaque "black boxes" and overlooks a critical question: is a diagnostic conclusion reached through a sound and verifiable reasoning pathway? The inscrutable nature of these systems poses a significant risk in high-stakes medical applications, potentially leading to flawed or untrustworthy conclusions. To address this, we conduct a large-scale empirical study of 3,600 cases from six medical datasets and six representative multi-agent frameworks. Through a rigorous, mixed-methods approach combining qualitative analysis with quantitative auditing, we develop a comprehensive taxonomy of collaborative failure modes. Our quantitative audit reveals four dominant failure patterns: flawed consensus driven by shared model deficiencies, suppression of correct minority opinions, ineffective discussion dynamics, and critical information loss during synthesis. This study demonstrates that high accuracy alone is an insufficient measure of clinical or public trust. It highlights the urgent need for transparent and auditable reasoning processes, a cornerstone for the responsible development and deployment of medical AI.

Huiya Zhao, Yinghao Zhu, Zixiang Wang et al.

The efficacy of AI agents in healthcare research is hindered by their reliance on static, predefined strategies. This creates a critical limitation: agents can become better tool-users but cannot learn to become better strategic planners, a crucial skill for complex domains like healthcare. We introduce HealthFlow, a self-evolving AI agent that overcomes this limitation through a novel meta-level evolution mechanism. HealthFlow autonomously refines its own high-level problem-solving policies by distilling procedural successes and failures into a durable, strategic knowledge base. To anchor our research and facilitate reproducible evaluation, we introduce EHRFlowBench, a new benchmark featuring complex, realistic health data analysis tasks derived from peer-reviewed clinical research. Our comprehensive experiments demonstrate that HealthFlow's self-evolving approach significantly outperforms state-of-the-art agent frameworks. This work marks a necessary shift from building better tool-users to designing smarter, self-evolving task-managers, paving the way for more autonomous and effective AI for scientific discovery.

Bo Hou, Xin Tan, Kai Zheng et al.

Atomic commits, which address a single development concern, are a best practice in software development. In practice, however, developers often produce tangled commits that mix unrelated changes, complicating code review and maintenance. Prior untangling approaches (rule-based, feature-based, or graph-based) have made progress but typically rely on shallow signals and struggle to distinguish explicit dependencies (e.g., control/data flow) from implicit ones (e.g., semantic or conceptual relationships). In this paper, we propose ColaUntangle, a new collaborative consultation framework for commit untangling that models both explicit and implicit dependencies among code changes. ColaUntangle integrates Large Language Model (LLM)-driven agents in a multi-agent architecture: one agent specializes in explicit dependencies, another in implicit ones, and a reviewer agent synthesizes their perspectives through iterative consultation. To capture structural and contextual information, we construct Explicit and Implicit Contexts, enabling agents to reason over code relationships with both symbolic and semantic depth. We evaluate ColaUntangle on two widely-used datasets (1,612 C# and 14k Java tangled commits). Experimental results show that ColaUntangle outperforms the best-performing baseline, achieving an improvement of 44% on the C# dataset and 82% on the Java dataset. These findings highlight the potential of LLM-based collaborative frameworks for advancing automated commit untangling tasks.

Shiyun Xie, Zhiru Wang, Yinghao Zhu et al.

Recently, 3D Gaussian Splatting (3DGS) has excelled in novel view synthesis (NVS) with its real-time rendering capabilities and superior quality. However, it encounters challenges for high-resolution novel view synthesis (HRNVS) due to the coarse nature of primitives derived from low-resolution input views. To address this issue, we propose SuperGS, an expansion of Scaffold-GS designed with a two-stage coarse-to-fine training framework. In the low-resolution stage, we introduce a latent feature field to represent the low-resolution scene, which serves as both the initialization and foundational information for super-resolution optimization. In the high-resolution stage, we propose a multi-view consistent densification strategy that backprojects high-resolution depth maps based on error maps and employs a multi-view voting mechanism, mitigating ambiguities caused by multi-view inconsistencies in the pseudo labels provided by 2D prior models while avoiding Gaussian redundancy. Furthermore, we model uncertainty through variational feature learning and use it to guide further scene representation refinement and adjust the supervisory effect of pseudo-labels, ensuring consistent and detailed scene reconstruction. Extensive experiments demonstrate that SuperGS outperforms state-of-the-art HRNVS methods on both forward-facing and 360-degree datasets.

Yinghao Zhu, Xiaochen Zheng, Ahmed Allam et al.

We propose TAMER, a Test-time Adaptive MoE-driven framework for Electronic Health Record (EHR) Representation learning. TAMER introduces a framework where a Mixture-of-Experts (MoE) architecture is co-designed with Test-Time Adaptation (TTA) to jointly mitigate the intertwined challenges of patient heterogeneity and distribution shifts in EHR modeling. The MoE focuses on latent patient subgroups through domain-aware expert specialization, while TTA enables real-time adaptation to evolving health status distributions when new patient samples are introduced. Extensive experiments across four real-world EHR datasets demonstrate that TAMER consistently improves predictive performance for both mortality and readmission risk tasks when combined with diverse EHR modeling backbones. TAMER offers a promising approach for dynamic and personalized EHR-based predictions in practical clinical settings.