Recommendation systems are key to boosting user engagement,satisfaction,and retention,particularly on media platforms where personalized content is vital.Sequential recommendation systems learn from user-item interact...Recommendation systems are key to boosting user engagement,satisfaction,and retention,particularly on media platforms where personalized content is vital.Sequential recommendation systems learn from user-item interactions to predict future items of interest.However,many current methods rely on unique user and item IDs,limiting their ability to represent users and items effectively,especially in zero-shot learning scenarios where training data is scarce.With the rapid development of Large Language Models(LLMs),researchers are exploring their potential to enhance recommendation systems.However,there is a semantic gap between the linguistic semantics of LLMs and the collaborative semantics of recommendation systems,where items are typically indexed by IDs.Moreover,most research focuses on item representations,neglecting personalized user modeling.To address these issues,we propose a sequential recommendation framework using LLMs,called CIT-Rec,a model that integrates Collaborative semantics for user representation and Image and Text information for item representation to enhance Recommendations.Specifically,by aligning intuitive image information with text containing semantic features,we can more accurately represent items,improving item representation quality.We focus not only on item representations but also on user representations.To more precisely capture users’personalized preferences,we use traditional sequential recommendation models to train on users’historical interaction data,effectively capturing behavioral patterns.Finally,by combining LLMs and traditional sequential recommendation models,we allow the LLM to understand linguistic semantics while capturing collaborative semantics.Extensive evaluations on real-world datasets show that our model outperforms baseline methods,effectively combining user interaction history with item visual and textual modalities to provide personalized recommendations.展开更多
Background:Assess ChatGPT and Bard's effectiveness in the initial identification of articles for Otolaryngology—Head and Neck Surgery systematic literature reviews.Methods:Three PRISMA-based systematic reviews(Ja...Background:Assess ChatGPT and Bard's effectiveness in the initial identification of articles for Otolaryngology—Head and Neck Surgery systematic literature reviews.Methods:Three PRISMA-based systematic reviews(Jabbour et al.2017,Wong et al.2018,and Wu et al.2021)were replicated using ChatGPTv3.5 and Bard.Outputs(author,title,publication year,and journal)were compared to the original references and cross-referenced with medical databases for authenticity and recall.Results:Several themes emerged when comparing Bard and ChatGPT across the three reviews.Bard generated more outputs and had greater recall in Wong et al.'s review,with a broader date range in Jabbour et al.'s review.In Wu et al.'s review,ChatGPT-2 had higher recall and identified more authentic outputs than Bard-2.Conclusion:Large language models(LLMs)failed to fully replicate peer-reviewed methodologies,producing outputs with inaccuracies but identifying relevant,especially recent,articles missed by the references.While human-led PRISMA-based reviews remain the gold standard,refining LLMs for literature reviews shows potential.展开更多
With direct expression of individual application domain patterns and ideas,domain-specific modeling language（DSML） is more and more frequently used to build models instead of using a combination of one or more gener...With direct expression of individual application domain patterns and ideas,domain-specific modeling language（DSML） is more and more frequently used to build models instead of using a combination of one or more general constructs.Based on the profile mechanism of unified modeling language（UML） 2.2,a kind of DSML is presented to model simulation testing systems of avionic software（STSAS）.To define the syntax,semantics and notions of the DSML,the domain model of the STSAS from which we generalize the domain concepts and relationships among these concepts is given,and then,the domain model is mapped into a UML meta-model,named UML-STSAS profile.Assuming a flight control system（FCS） as system under test（SUT）,we design the relevant STSAS.The results indicate that extending UML to the simulation testing domain can effectively and precisely model STSAS.展开更多
Data-driven approaches are extensively employed to model complex chemical engineering processes, such as hydrotreating, to address the challenges of mechanism-based methods demanding deep process understanding. Howeve...Data-driven approaches are extensively employed to model complex chemical engineering processes, such as hydrotreating, to address the challenges of mechanism-based methods demanding deep process understanding. However, the development of such models requires specialized expertise in data science, limiting their broader application. Large language models (LLMs), such as GPT-4, have demonstrated potential in supporting and guiding research efforts. This work presents a novel AI-assisted framework where GPT-4, through well-engineered prompts, facilitates the construction and explanation of multi-objective neural networks. These models predict hydrotreating products properties (such as distillation range), including refined diesel and refined gas oil, using feedstock properties, operating conditions, and recycle hydrogen composition. Gradient-weighted class activation mapping was employed to identify key features influencing the output variables. This work illustrates an innovative AI-guided paradigm for chemical engineering applications, and the designed prompts hold promise for adaptation to other complex processes.展开更多
Dear Editor,This letter deals with automatically constructing an OPC UA information model(IM)aimed at enhancing data interoperability among heterogeneous system components within manufacturing automation systems.Empow...Dear Editor,This letter deals with automatically constructing an OPC UA information model(IM)aimed at enhancing data interoperability among heterogeneous system components within manufacturing automation systems.Empowered by the large language model(LLM),we propose a novel multi-agent collaborative framework to streamline the end-to-end OPC UA IM modeling process.Each agent is equipped with meticulously engineered prompt templates,augmenting their capacity to execute specific tasks.We conduct modeling experiments using real textual data to demonstrate the effectiveness of the proposed method,improving modeling efficiency and reducing the labor workload.展开更多
Recent advancements in large language models(LLMs)have driven remarkable progress in text process-ing,opening new avenues for medical knowledge discovery.In this study,we present ERQA,a mEdical knowledge Retrieval and...Recent advancements in large language models(LLMs)have driven remarkable progress in text process-ing,opening new avenues for medical knowledge discovery.In this study,we present ERQA,a mEdical knowledge Retrieval and Question-Answering framework powered by an enhanced LLM that integrates a semantic vector database and a curated literature repository.The ERQA framework leverages domain-specific incremental pretraining and conducts supervised fine-tuning on medical literature,enabling retrieval and question-answering(QA)tasks to be completed with high precision.Performance evaluations implemented on the coronavirus disease 2019(COVID-19)and TripClick data-sets demonstrate the robust capabilities of ERQA across multiple tasks.On the COVID-19 dataset,ERQA-13B achieves state-of-the-art retrieval metrics,with normalized discounted cumulative gain at top 10(NDCG@10)0.297,recall values at top 10(Recall@10)0.347,and mean reciprocal rank(MRR)=0.370;it also attains strong abstract summarization performance,with a recall-oriented understudy for gisting evaluation(ROUGE)-1 score of 0.434,and QA performance,with a bilingual evaluation understudy(BLEU)-1 score of 7.851.The comparable performance achieved on the TripClick dataset further under-scores the adaptability of ERQA across diverse medical topics.These findings suggest that ERQA repre-sents a significant step toward efficient biomedical knowledge retrieval and QA.展开更多
Short Message Service(SMS)is a widely used and cost-effective communication medium that has unfortunately become a frequent target for unsolicited messages-commonly known as SMS spam.With the rapid adoption of smartph...Short Message Service(SMS)is a widely used and cost-effective communication medium that has unfortunately become a frequent target for unsolicited messages-commonly known as SMS spam.With the rapid adoption of smartphones and increased Internet connectivity,SMS spam has emerged as a prevalent threat.Spammers have recognized the critical role SMS plays in today’s modern communication,making it a prime target for abuse.As cybersecurity threats continue to evolve,the volume of SMS spam has increased substantially in recent years.Moreover,the unstructured format of SMS data creates significant challenges for SMS spam detection,making it more difficult to successfully combat spam attacks.In this paper,we present an optimized and fine-tuned transformer-based Language Model to address the problem of SMS spam detection.We use a benchmark SMS spam dataset to analyze this spam detection model.Additionally,we utilize pre-processing techniques to obtain clean and noise-free data and address class imbalance problem by leveraging text augmentation techniques.The overall experiment showed that our optimized fine-tuned BERT(Bidirectional Encoder Representations from Transformers)variant model RoBERTa obtained high accuracy with 99.84%.To further enhance model transparency,we incorporate Explainable Artificial Intelligence(XAI)techniques that compute positive and negative coefficient scores,offering insight into the model’s decision-making process.Additionally,we evaluate the performance of traditional machine learning models as a baseline for comparison.This comprehensive analysis demonstrates the significant impact language models can have on addressing complex text-based challenges within the cybersecurity landscape.展开更多
Objective:Generative artificial intelligence(AI)technology,represented by large language models(LLMs),has gradually been developed for traditional Chinese medicine(TCM);however,challenges remain in effectively enhanci...Objective:Generative artificial intelligence(AI)technology,represented by large language models(LLMs),has gradually been developed for traditional Chinese medicine(TCM);however,challenges remain in effectively enhancing AI applications for TCM.Therefore,this study is the first systematic review to analyze LLMs in TCM retrospectively,focusing on and summarizing the evidence of their performance in generative tasks.Methods:We extensively searched electronic databases for articles published until June 2024 to identify publicly available studies on LLMs in TCM.Two investigators independently selected and extracted the related information and evaluation metrics.Based on the available data,this study used descriptive analysis for a comprehensive systematic review of LLM technology related to TCM.Results:Ten studies published between 2023 and 2024 met our eligibility criteria and were included in this review,including 40%LLMs in the TCM vertical domain,40%containing TCM data,and 20%honoring the TCM contribution,with a foundational model parameter range from 1.8 to 33 billion.All included studies used manual or automatic evaluation metrics to evaluate model performance and fully discussed the challenges and contributions through an overview of LLMs in TCM.Conclusions:LLMs have achieved significant advantages in TCM applications and can effectively address intelligent TCM tasks.Further in-depth development of LLMs is needed in various vertical TCM fields,including clinical and fundamental research.Focusing on the functional segmentation development direction of generative AI technologies in TCM application scenarios to meet the practical needs-oriented demands of TCM digitalization is essential.展开更多
In the era of big data,data-driven technologies are increasingly leveraged by industry to facilitate autonomous learning and intelligent decision-making.However,the challenge of“small samples in big data”emerges whe...In the era of big data,data-driven technologies are increasingly leveraged by industry to facilitate autonomous learning and intelligent decision-making.However,the challenge of“small samples in big data”emerges when datasets lack the comprehensive information necessary for addressing complex scenarios,which hampers adaptability.Thus,enhancing data completeness is essential.Knowledge-guided virtual sample generation transforms domain knowledge into extensive virtual datasets,thereby reducing dependence on limited real samples and enabling zero-sample fault diagnosis.This study used building air conditioning systems as a case study.We innovatively used the large language model(LLM)to acquire domain knowledge for sample generation,significantly lowering knowledge acquisition costs and establishing a generalized framework for knowledge acquisition in engineering applications.This acquired knowledge guided the design of diffusion boundaries in mega-trend diffusion(MTD),while the Monte Carlo method was used to sample within the diffusion function to create information-rich virtual samples.Additionally,a noise-adding technique was introduced to enhance the information entropy of these samples,thereby improving the robustness of neural networks trained with them.Experimental results showed that training the diagnostic model exclusively with virtual samples achieved an accuracy of 72.80%,significantly surpassing traditional small-sample supervised learning in terms of generalization.This underscores the quality and completeness of the generated virtual samples.展开更多
Model evaluation using benchmark datasets is an important method to measure the capability of large language models(LLMs)in specific domains,and it is mainly used to assess the knowledge and reasoning abilities of LLM...Model evaluation using benchmark datasets is an important method to measure the capability of large language models(LLMs)in specific domains,and it is mainly used to assess the knowledge and reasoning abilities of LLMs.Therefore,in order to better assess the capability of LLMs in the agricultural domain,Agri-Eval was proposed as a benchmark for assessing the knowledge and reasoning ability of LLMs in agriculture.The assessment dataset used in Agri-Eval covered seven major disciplines in the agricultural domain:crop science,horticulture,plant protection,animal husbandry,forest science,aquaculture science,and grass science,and contained a total of 2283 questions.Among domestic general-purpose LLMs,DeepSeek R1 performed best with an accuracy rate of 75.49%.In the realm of international general-purpose LLMs,Gemini 2.0 pro exp 0205 standed out as the top performer,achieving an accuracy rate of 74.28%.As an LLMs in agriculture vertical,Shennong V2.0 outperformed all the LLMs in China,and the answer accuracy rate of agricultural knowledge exceeded that of all the existing general-purpose LLMs.The launch of Agri-Eval helped the LLM developers to comprehensively evaluate the model's capability in the field of agriculture through a variety of tasks and tests to promote the development of the LLMs in the field of agriculture.展开更多
Background:Large language models(LLMs)have shown considerable promise in supporting clinical decision-making.However,their adoption and evaluation in dermatology remains limited.This study aimed to explore the prefere...Background:Large language models(LLMs)have shown considerable promise in supporting clinical decision-making.However,their adoption and evaluation in dermatology remains limited.This study aimed to explore the preferences of Chinese dermatologists regarding LLM-generated responses in clinical psoriasis scenarios and to assess how they prioritize key quality dimensions,including accuracy,traceability,and logicality.Methods:A cross-sectional,web-based survey was conducted between December 25,2024,and January 22,2025,following the Checklist for Reporting Results of Internet E-Surveys guidelines.A total of 1247 valid responses were collected from practicing dermatologists across 33 of China's provincial-level administrative divisions.Participants evaluated responses to five categories of clinical questions(etiology,clinical presentation,differential diagnosis,treatment,and case study)generated by five LLMs:ChatGPT-4o,Kimi.ai,Doubao,ZuoYiGPT,and Lingyi-agent.Statistical associations between participant characteristics and model preferences were examined using chi-square tests.Results:ChatGPT-4o(Model 1)emerged as the most preferred model across all clinical tasks,consistently receiving the highest number of votes in case study(n=740),clinical presentation(n=666),differential diagnosis(n=707),etiology(n=602),and treatment(n=656).Significant variation in model preference by professional title was observed only for the differential diagnosis task(χ^(2)=21.13,df=12,p=0.0485),while no significant differences were found across hospital tiers(p>0.05).In terms of evaluation dimensions,accuracy was most frequently rated as“very important”(n=635).A significant association existed between hospital tier and the most valued dimension(χ^(2)=27.667,df=9,p=0.0011),with dermatologists in primary hospitals prioritizing traceability more than their peers in higher-tier hospitals.No significant associations were found across professional titles(p=0.127).Conclusions:Chinese dermatologists suggest a strong preference for ChatGPT-4o over domestic LLMs in psoriasis-related clinical tasks.While accuracy remains the primary criterion,traceability and logicality are also critical,particularly for clinicians in lower-tier hospitals.These findings suggest that future clinical LLMs should prioritize not only content accuracy but also source transparency and structural clarity to meet the diverse needs of different clinical settings.展开更多
The collection and annotation of lar ge-scale bird datasets are resource-intensive and time-consuming processes that significantly limit the scalability and accuracy of biodiversity monitoring systems.While self-super...The collection and annotation of lar ge-scale bird datasets are resource-intensive and time-consuming processes that significantly limit the scalability and accuracy of biodiversity monitoring systems.While self-supervised learning(SSL)has emerged as a promising approach for leveraging unannotated data,current SSL methods face two critical challenges in bird species recognition:(1)long-tailed data distributions that result in poor performance on underrepresented species;and(2)domain shift issues caused by data augmentation strategies designed to mitigate class imbalance.Here we present SDNet,a novel SSL-based bird recognition framework that integrates diffusion models with large language models(LLMs)to overcome these limitations.SDNet employs LLMs to generate semantically rich textual descriptions for tail-class species by prompting the models with species taxonomy,morphological attributes,and habitat information,producing detailed natural language priors that capture fine-grained visual characteristics(e.g.,plumage patterns,body proportions,and distinctive markings).These textual descriptions are subsequently used by a conditional diffusion model to synthesize new bird image samples through cross-attention mechanisms that fuse textual embeddings with intermediate visual feature representations during the denoising process,ensuring generated images preserve species-specific morphological details while maintaining photorealistic quality.Additionally,we incorporate a Swin Transformer as the feature extraction backbone whose hierarchical window-based attention mechanism and shifted windowing scheme enable multi-scale local feature extraction that proves particularly effective at capturing finegrained discriminative patterns(such as beak shape and feather texture)while mitigating domain shift between synthetic and original images through consistent feature representations across both data sources.SDNet is validated on both a self-constructed dataset(Bird_BXS)an d a publicly available benchmark(Birds_25),demonstrating substantial improvements over conventional SSL approaches.Our results indicate that the synergistic integration of LLMs,diffusion models,and the Swin Transformer architecture contributes significantly to recognition accuracy,particularly for rare and morphologically similar species.These findings highlight the potential of SDNet for addressing fundamental limitations of existing SSL methods in avian recognition tasks and establishing a new paradigm for efficient self-supervised learning in large-scale ornithological vision applications.展开更多
This study demonstrates a novel integration of large language models,machine learning,and multicriteria decision-making to investigate self-moderation in small online communities,a topic under-explored compared to use...This study demonstrates a novel integration of large language models,machine learning,and multicriteria decision-making to investigate self-moderation in small online communities,a topic under-explored compared to user behavior and platform-driven moderation on social media.The proposed methodological framework(1)utilizes large language models for social media post analysis and categorization,(2)employs k-means clustering for content characterization,and(3)incorporates the TODIM(Tomada de Decisão Interativa Multicritério)method to determine moderation strategies based on expert judgments.In general,the fully integrated framework leverages the strengths of these intelligent systems in a more systematic evaluation of large-scale decision problems.When applied in social media moderation,this approach promotes nuanced and context-sensitive self-moderation by taking into account factors such as cultural background and geographic location.The application of this framework is demonstrated within Facebook groups.Eight distinct content clusters encompassing safety,harassment,diversity,and misinformation are identified.Analysis revealed a preference for content removal across all clusters,suggesting a cautious approach towards potentially harmful content.However,the framework also highlights the use of other moderation actions,like account suspension,depending on the content category.These findings contribute to the growing body of research on self-moderation and offer valuable insights for creating safer and more inclusive online spaces within smaller communities.展开更多
Objective To develop a clinical decision and prescription generation system(CDPGS)specifically for diarrhea in traditional Chinese medicine(TCM),utilizing a specialized large language model(LLM),Qwen-TCM-Dia,to standa...Objective To develop a clinical decision and prescription generation system(CDPGS)specifically for diarrhea in traditional Chinese medicine(TCM),utilizing a specialized large language model(LLM),Qwen-TCM-Dia,to standardize diagnostic processes and prescription generation.Methods Two primary datasets were constructed:an evaluation benchmark and a fine-tuning dataset consisting of fundamental diarrhea knowledge,medical records,and chain-ofthought(CoT)reasoning datasets.After an initial evaluation of 16 open-source LLMs across inference time,accuracy,and output quality,Qwen2.5 was selected as the base model due to its superior overall performance.We then employed a two-stage low-rank adaptation(LoRA)fine-tuning strategy,integrating continued pre-training on domain-specific knowledge with instruction fine-tuning using CoT-enriched medical records.This approach was designed to embed the clinical logic(symptoms→pathogenesis→therapeutic principles→prescriptions)into the model’s reasoning capabilities.The resulting fine-tuned model,specialized for TCM diarrhea,was designated as Qwen-TCM-Dia.Model performance was evaluated for disease diagnosis and syndrome type differentiation using accuracy,precision,recall,and F1-score.Furthermore,the quality of the generated prescriptions was compared with that of established open-source TCM LLMs.Results Qwen-TCM-Dia achieved peak performance compared to both the base Qwen2.5 model and five other open-source TCM LLMs.It achieved 97.05%accuracy and 91.48%F1-score in disease diagnosis,and 74.54%accuracy and 74.21%F1-score in syndrome type differentiation.Compared with existing open-source TCM LLMs(BianCang,HuangDi,LingDan,TCMLLM-PR,and ZhongJing),Qwen-TCM-Dia exhibited higher fidelity in reconstructing the“symptoms→pathogenesis→therapeutic principles→prescriptions”logic chain.It provided complete prescriptions,whereas other models often omitted dosages or generated mismatched prescriptions.Conclusion By integrating continued pre-training,CoT reasoning,and a two-stage fine-tuning strategy,this study establishes a CDPGS for diarrhea in TCM.The results demonstrate the synergistic effect of strengthening domain representation through pre-training and activating logical reasoning via CoT.This research not only provides critical technical support for the standardized diagnosis and treatment of diarrhea but also offers a scalable paradigm for the digital inheritance of expert TCM experience and the intelligent transformation of TCM.展开更多
The use of Unmanned Aerial Vehicles(UAVs)for defect detection on railway slopes is becoming increasingly widespread due to their ability to capture high-resolution images over large,inaccessible,and topographically co...The use of Unmanned Aerial Vehicles(UAVs)for defect detection on railway slopes is becoming increasingly widespread due to their ability to capture high-resolution images over large,inaccessible,and topographically complex areas.However,current UAV-based detection methods face several critical limitations,including constrained deployment frequency,limited availability of annotated defect data,and the lack of mature risk assessment frameworks.To address these challenges,this study introduces a novel approach that integrates diffusion models with Large Language Models(LLMs)to generate highquality synthetic defect images tailored to railway slope scenarios.Furthermore,an improved transformerbased architecture is proposed,incorporating attention mechanisms and LLM-guided diffusion-generated imagery to enhance defect recognition performance under complex environmental conditions.Experimental evaluations conducted on a dataset of 300 field-collected images from high-risk railway slopes demonstrate that the proposed method significantly outperforms existing baselines in terms of precision,recall,and robustness,indicating strong applicability for real-world railway infrastructure monitoring and disaster prevention.展开更多
Classifying job offers into occupational categories is a fundamental task in human resource information systems,as it improves and streamlines indexing,search,and matching between openings and job seekers.Comprehensiv...Classifying job offers into occupational categories is a fundamental task in human resource information systems,as it improves and streamlines indexing,search,and matching between openings and job seekers.Comprehensive occupational databases such as O∗NET or ESCO provide detailed taxonomies of interrelated positions that can be leveraged to align the textual content of postings with occupational categories,thereby facilitating standardization,cross-system interoperability,and access to metadata for each occupation(e.g.,tasks,knowledge,skills,and abilities).In this work,we explore the effectiveness of fine-tuning existing language models(LMs)to classify job offers with occupational descriptors from O∗NET.This enables a more precise assessment of candidate suitability by identifying the specific knowledge and skills required for each position,and helps automate recruitment processes by mitigating human bias and subjectivity in candidate selection.We evaluate three representative BERT-like models:BERT,RoBERTa,and DeBERTa.BERT serves as the baseline encoder-only architecture;RoBERTa incorporates advances in pretraining objectives and data scale;and DeBERTa introduces architectural improvements through disentangled attention mechanisms.The best performance was achieved with the DeBERTa model,although the other models also produced strong results,and no statistically significant differences were observed acrossmodels.We also find that these models typically reach optimal performance after only a few training epochs,and that training with smaller,balanced datasets is effective.Consequently,comparable results can be obtained with models that require fewer computational resources and less training time,facilitating deployment and practical use.展开更多
Knowledge distillation has become a standard technique for compressing large language models into efficient student models,but existing methods often struggle to balance prediction accuracy with explanation quality.Re...Knowledge distillation has become a standard technique for compressing large language models into efficient student models,but existing methods often struggle to balance prediction accuracy with explanation quality.Recent approaches such as Distilling Step-by-Step(DSbS)introduce explanation supervision,yet they apply it in a uniform manner that may not fully exploit the different learning dynamics of prediction and explanation.In this work,we propose a task-structured curriculum learning(TSCL)framework that structures training into three sequential phases:(i)prediction-only,to establish stable feature representations;(ii)joint prediction-explanation,to align task outputs with rationale generation;and(iii)explanation-only,to refine the quality of rationales.This design provides a simple but effective modification to DSbS,requiring no architectural changes and adding negligible training cost.We justify the phase scheduling with ablation studies and convergence analysis,showing that an initial prediction-heavy stage followed by a balanced joint phase improves both stability and explanation alignment.Extensive experiments on five datasets(e-SNLI,ANLI,CommonsenseQA,SVAMP,and MedNLI)demonstrate that TSCL consistently outperforms strong baselines,achieving gains of+1.7-2.6 points in accuracy and 0.8-1.2 in ROUGE-L,corresponding to relative error reductions of up to 21%.Beyond lexical metrics,human evaluation and ERASERstyle faithfulness diagnostics confirm that TSCL produces more faithful and informative explanations.Comparative training curves further reveal faster convergence and lower variance across seeds.Efficiency analysis shows less than 3%overhead in wall-clock training time and no additional inference cost,making the approach practical for realworld deployment.This study demonstrates that a simple task-structured curriculum can significantly improve the effectiveness of knowledge distillation.By separating and sequencing objectives,TSCL achieves a better balance between accuracy,stability,and explanation quality.The framework generalizes across domains,including medical NLI,and offers a principled recipe for future applications in multimodal reasoning and reinforcement learning.展开更多
The malicious dissemination of hate speech via compromised accounts,automated bot networks and malware-driven social media campaigns has become a growing cybersecurity concern.Automatically detecting such content in S...The malicious dissemination of hate speech via compromised accounts,automated bot networks and malware-driven social media campaigns has become a growing cybersecurity concern.Automatically detecting such content in Spanish is challenging due to linguistic complexity and the scarcity of annotated resources.In this paper,we compare two predominant AI-based approaches for the forensic detection of malicious hate speech:(1)finetuning encoder-only models that have been trained in Spanish and(2)In-Context Learning techniques(Zero-and Few-Shot Learning)with large-scale language models.Our approach goes beyond binary classification,proposing a comprehensive,multidimensional evaluation that labels each text by:(1)type of speech,(2)recipient,(3)level of intensity(ordinal)and(4)targeted group(multi-label).Performance is evaluated using an annotated Spanish corpus,standard metrics such as precision,recall and F1-score and stability-oriented metrics to evaluate the stability of the transition from zero-shot to few-shot prompting(Zero-to-Few Shot Retention and Zero-to-Few Shot Gain)are applied.The results indicate that fine-tuned encoder-only models(notably MarIA and BETO variants)consistently deliver the strongest and most reliable performance:in our experiments their macro F1-scores lie roughly in the range of approximately 46%–66%depending on the task.Zero-shot approaches are much less stable and typically yield substantially lower performance(observed F1-scores range approximately 0%–39%),often producing invalid outputs in practice.Few-shot prompting(e.g.,Qwen 38B,Mistral 7B)generally improves stability and recall relative to pure zero-shot,bringing F1-scores into a moderate range of approximately 20%–51%but still falling short of fully fine-tuned models.These findings highlight the importance of supervised adaptation and discuss the potential of both paradigms as components in AI-powered cybersecurity and malware forensics systems designed to identify and mitigate coordinated online hate campaigns.展开更多
Large language models(LLMs)show considerable potential to revolutionize healthcare through their performance across diverse clinical applications.Given the inherent constraints of LLMs and the critical nature of medic...Large language models(LLMs)show considerable potential to revolutionize healthcare through their performance across diverse clinical applications.Given the inherent constraints of LLMs and the critical nature of medical practice,a rigorous and systematic evaluation of their medical competence is imperative.This study presents a comprehensive review of the established methodologies and benchmarks for evaluating the medical competence of LLMs,encompassing a thorough analysis of current assessment practices across medical knowledge,clinical practice competence,and ethical-safety considerations.By integrating clinician competency assessment frameworks into LLMs evaluation,we propose a structured tri-dimensional framework that systematically organizes existing evaluation approaches according to medical theoretical knowledge,clinical practice ability,and ethical-safety considerations.Furthermore,this research provides critical insights into future developmental trajectories while establishing foundational frameworks and standardization protocols for the integration of LLMs into medical practice.展开更多
基金supported by the National Key R&D Program of China[2022YFF0902703]the State Administration for Market Regulation Science and Technology Plan Project(2024MK033).
文摘Recommendation systems are key to boosting user engagement,satisfaction,and retention,particularly on media platforms where personalized content is vital.Sequential recommendation systems learn from user-item interactions to predict future items of interest.However,many current methods rely on unique user and item IDs,limiting their ability to represent users and items effectively,especially in zero-shot learning scenarios where training data is scarce.With the rapid development of Large Language Models(LLMs),researchers are exploring their potential to enhance recommendation systems.However,there is a semantic gap between the linguistic semantics of LLMs and the collaborative semantics of recommendation systems,where items are typically indexed by IDs.Moreover,most research focuses on item representations,neglecting personalized user modeling.To address these issues,we propose a sequential recommendation framework using LLMs,called CIT-Rec,a model that integrates Collaborative semantics for user representation and Image and Text information for item representation to enhance Recommendations.Specifically,by aligning intuitive image information with text containing semantic features,we can more accurately represent items,improving item representation quality.We focus not only on item representations but also on user representations.To more precisely capture users’personalized preferences,we use traditional sequential recommendation models to train on users’historical interaction data,effectively capturing behavioral patterns.Finally,by combining LLMs and traditional sequential recommendation models,we allow the LLM to understand linguistic semantics while capturing collaborative semantics.Extensive evaluations on real-world datasets show that our model outperforms baseline methods,effectively combining user interaction history with item visual and textual modalities to provide personalized recommendations.
文摘Background:Assess ChatGPT and Bard's effectiveness in the initial identification of articles for Otolaryngology—Head and Neck Surgery systematic literature reviews.Methods:Three PRISMA-based systematic reviews(Jabbour et al.2017,Wong et al.2018,and Wu et al.2021)were replicated using ChatGPTv3.5 and Bard.Outputs(author,title,publication year,and journal)were compared to the original references and cross-referenced with medical databases for authenticity and recall.Results:Several themes emerged when comparing Bard and ChatGPT across the three reviews.Bard generated more outputs and had greater recall in Wong et al.'s review,with a broader date range in Jabbour et al.'s review.In Wu et al.'s review,ChatGPT-2 had higher recall and identified more authentic outputs than Bard-2.Conclusion:Large language models(LLMs)failed to fully replicate peer-reviewed methodologies,producing outputs with inaccuracies but identifying relevant,especially recent,articles missed by the references.While human-led PRISMA-based reviews remain the gold standard,refining LLMs for literature reviews shows potential.
基金Aeronautical Science Foundation of China (20095551025)
文摘With direct expression of individual application domain patterns and ideas,domain-specific modeling language（DSML） is more and more frequently used to build models instead of using a combination of one or more general constructs.Based on the profile mechanism of unified modeling language（UML） 2.2,a kind of DSML is presented to model simulation testing systems of avionic software（STSAS）.To define the syntax,semantics and notions of the DSML,the domain model of the STSAS from which we generalize the domain concepts and relationships among these concepts is given,and then,the domain model is mapped into a UML meta-model,named UML-STSAS profile.Assuming a flight control system（FCS） as system under test（SUT）,we design the relevant STSAS.The results indicate that extending UML to the simulation testing domain can effectively and precisely model STSAS.
基金supported by the National Key Research and Development Program of China(2023YFA1507601)the National Natural Science Foundation of China(22278127,22378038)+2 种基金the Fundamental Research Funds for the Central Universities(2022ZFJH004)the Shanghai Pilot Program for Basic Research(22T01400100-18)the Natural Science Foundation of Liaoning Province,China(2024-MSBA-15).
文摘Data-driven approaches are extensively employed to model complex chemical engineering processes, such as hydrotreating, to address the challenges of mechanism-based methods demanding deep process understanding. However, the development of such models requires specialized expertise in data science, limiting their broader application. Large language models (LLMs), such as GPT-4, have demonstrated potential in supporting and guiding research efforts. This work presents a novel AI-assisted framework where GPT-4, through well-engineered prompts, facilitates the construction and explanation of multi-objective neural networks. These models predict hydrotreating products properties (such as distillation range), including refined diesel and refined gas oil, using feedstock properties, operating conditions, and recycle hydrogen composition. Gradient-weighted class activation mapping was employed to identify key features influencing the output variables. This work illustrates an innovative AI-guided paradigm for chemical engineering applications, and the designed prompts hold promise for adaptation to other complex processes.
基金supported supported by the Fundamental Research Funds for the Central Universities(226-2024-00004)the National Natural Science Foundation of China(U23 A20326)Key Research and Development Program of Zhejiang Province(2025C01061).
文摘Dear Editor,This letter deals with automatically constructing an OPC UA information model(IM)aimed at enhancing data interoperability among heterogeneous system components within manufacturing automation systems.Empowered by the large language model(LLM),we propose a novel multi-agent collaborative framework to streamline the end-to-end OPC UA IM modeling process.Each agent is equipped with meticulously engineered prompt templates,augmenting their capacity to execute specific tasks.We conduct modeling experiments using real textual data to demonstrate the effectiveness of the proposed method,improving modeling efficiency and reducing the labor workload.
基金supported by the Innovation Fund for Medical Sciences of the Chinese Academy of Medical Sciences(2021-I2M-1-033)the National Key Research and Development Program of China(2022YFF0711900).
文摘Recent advancements in large language models(LLMs)have driven remarkable progress in text process-ing,opening new avenues for medical knowledge discovery.In this study,we present ERQA,a mEdical knowledge Retrieval and Question-Answering framework powered by an enhanced LLM that integrates a semantic vector database and a curated literature repository.The ERQA framework leverages domain-specific incremental pretraining and conducts supervised fine-tuning on medical literature,enabling retrieval and question-answering(QA)tasks to be completed with high precision.Performance evaluations implemented on the coronavirus disease 2019(COVID-19)and TripClick data-sets demonstrate the robust capabilities of ERQA across multiple tasks.On the COVID-19 dataset,ERQA-13B achieves state-of-the-art retrieval metrics,with normalized discounted cumulative gain at top 10(NDCG@10)0.297,recall values at top 10(Recall@10)0.347,and mean reciprocal rank(MRR)=0.370;it also attains strong abstract summarization performance,with a recall-oriented understudy for gisting evaluation(ROUGE)-1 score of 0.434,and QA performance,with a bilingual evaluation understudy(BLEU)-1 score of 7.851.The comparable performance achieved on the TripClick dataset further under-scores the adaptability of ERQA across diverse medical topics.These findings suggest that ERQA repre-sents a significant step toward efficient biomedical knowledge retrieval and QA.
文摘Short Message Service(SMS)is a widely used and cost-effective communication medium that has unfortunately become a frequent target for unsolicited messages-commonly known as SMS spam.With the rapid adoption of smartphones and increased Internet connectivity,SMS spam has emerged as a prevalent threat.Spammers have recognized the critical role SMS plays in today’s modern communication,making it a prime target for abuse.As cybersecurity threats continue to evolve,the volume of SMS spam has increased substantially in recent years.Moreover,the unstructured format of SMS data creates significant challenges for SMS spam detection,making it more difficult to successfully combat spam attacks.In this paper,we present an optimized and fine-tuned transformer-based Language Model to address the problem of SMS spam detection.We use a benchmark SMS spam dataset to analyze this spam detection model.Additionally,we utilize pre-processing techniques to obtain clean and noise-free data and address class imbalance problem by leveraging text augmentation techniques.The overall experiment showed that our optimized fine-tuned BERT(Bidirectional Encoder Representations from Transformers)variant model RoBERTa obtained high accuracy with 99.84%.To further enhance model transparency,we incorporate Explainable Artificial Intelligence(XAI)techniques that compute positive and negative coefficient scores,offering insight into the model’s decision-making process.Additionally,we evaluate the performance of traditional machine learning models as a baseline for comparison.This comprehensive analysis demonstrates the significant impact language models can have on addressing complex text-based challenges within the cybersecurity landscape.
基金supported by the National Multidisciplinary Innovation Team of Traditional Chinese Medicine(ZYYCXTD-D-202204)China Postdoctoral Science Foundation(2023M742627)+1 种基金Postdoctoral Fellowship Program of CPSF(GZC20231928)Foundation of State Key Laboratory of Component-based Chinese Medicine(CBCM2023201).
文摘Objective:Generative artificial intelligence(AI)technology,represented by large language models(LLMs),has gradually been developed for traditional Chinese medicine(TCM);however,challenges remain in effectively enhancing AI applications for TCM.Therefore,this study is the first systematic review to analyze LLMs in TCM retrospectively,focusing on and summarizing the evidence of their performance in generative tasks.Methods:We extensively searched electronic databases for articles published until June 2024 to identify publicly available studies on LLMs in TCM.Two investigators independently selected and extracted the related information and evaluation metrics.Based on the available data,this study used descriptive analysis for a comprehensive systematic review of LLM technology related to TCM.Results:Ten studies published between 2023 and 2024 met our eligibility criteria and were included in this review,including 40%LLMs in the TCM vertical domain,40%containing TCM data,and 20%honoring the TCM contribution,with a foundational model parameter range from 1.8 to 33 billion.All included studies used manual or automatic evaluation metrics to evaluate model performance and fully discussed the challenges and contributions through an overview of LLMs in TCM.Conclusions:LLMs have achieved significant advantages in TCM applications and can effectively address intelligent TCM tasks.Further in-depth development of LLMs is needed in various vertical TCM fields,including clinical and fundamental research.Focusing on the functional segmentation development direction of generative AI technologies in TCM application scenarios to meet the practical needs-oriented demands of TCM digitalization is essential.
基金supported by the National Natural Science Foundation of China(No.62306281)the Natural Science Foundation of Zhejiang Province(Nos.LQ23E060006 and LTGG24E050005)the Key Research Plan of Jiaxing City(No.2024BZ20016).
文摘In the era of big data,data-driven technologies are increasingly leveraged by industry to facilitate autonomous learning and intelligent decision-making.However,the challenge of“small samples in big data”emerges when datasets lack the comprehensive information necessary for addressing complex scenarios,which hampers adaptability.Thus,enhancing data completeness is essential.Knowledge-guided virtual sample generation transforms domain knowledge into extensive virtual datasets,thereby reducing dependence on limited real samples and enabling zero-sample fault diagnosis.This study used building air conditioning systems as a case study.We innovatively used the large language model(LLM)to acquire domain knowledge for sample generation,significantly lowering knowledge acquisition costs and establishing a generalized framework for knowledge acquisition in engineering applications.This acquired knowledge guided the design of diffusion boundaries in mega-trend diffusion(MTD),while the Monte Carlo method was used to sample within the diffusion function to create information-rich virtual samples.Additionally,a noise-adding technique was introduced to enhance the information entropy of these samples,thereby improving the robustness of neural networks trained with them.Experimental results showed that training the diagnostic model exclusively with virtual samples achieved an accuracy of 72.80%,significantly surpassing traditional small-sample supervised learning in terms of generalization.This underscores the quality and completeness of the generated virtual samples.
文摘Model evaluation using benchmark datasets is an important method to measure the capability of large language models(LLMs)in specific domains,and it is mainly used to assess the knowledge and reasoning abilities of LLMs.Therefore,in order to better assess the capability of LLMs in the agricultural domain,Agri-Eval was proposed as a benchmark for assessing the knowledge and reasoning ability of LLMs in agriculture.The assessment dataset used in Agri-Eval covered seven major disciplines in the agricultural domain:crop science,horticulture,plant protection,animal husbandry,forest science,aquaculture science,and grass science,and contained a total of 2283 questions.Among domestic general-purpose LLMs,DeepSeek R1 performed best with an accuracy rate of 75.49%.In the realm of international general-purpose LLMs,Gemini 2.0 pro exp 0205 standed out as the top performer,achieving an accuracy rate of 74.28%.As an LLMs in agriculture vertical,Shennong V2.0 outperformed all the LLMs in China,and the answer accuracy rate of agricultural knowledge exceeded that of all the existing general-purpose LLMs.The launch of Agri-Eval helped the LLM developers to comprehensively evaluate the model's capability in the field of agriculture through a variety of tasks and tests to promote the development of the LLMs in the field of agriculture.
基金National Key Research and Development Program of China,Grant/Award Number:2024YFF0507404Special Clinical Business Fund for High-Level Hospitals of China-Japan Friendship Hospital,Grant/Award Number:2024-NHLHCRF-TS-01。
文摘Background:Large language models(LLMs)have shown considerable promise in supporting clinical decision-making.However,their adoption and evaluation in dermatology remains limited.This study aimed to explore the preferences of Chinese dermatologists regarding LLM-generated responses in clinical psoriasis scenarios and to assess how they prioritize key quality dimensions,including accuracy,traceability,and logicality.Methods:A cross-sectional,web-based survey was conducted between December 25,2024,and January 22,2025,following the Checklist for Reporting Results of Internet E-Surveys guidelines.A total of 1247 valid responses were collected from practicing dermatologists across 33 of China's provincial-level administrative divisions.Participants evaluated responses to five categories of clinical questions(etiology,clinical presentation,differential diagnosis,treatment,and case study)generated by five LLMs:ChatGPT-4o,Kimi.ai,Doubao,ZuoYiGPT,and Lingyi-agent.Statistical associations between participant characteristics and model preferences were examined using chi-square tests.Results:ChatGPT-4o(Model 1)emerged as the most preferred model across all clinical tasks,consistently receiving the highest number of votes in case study(n=740),clinical presentation(n=666),differential diagnosis(n=707),etiology(n=602),and treatment(n=656).Significant variation in model preference by professional title was observed only for the differential diagnosis task(χ^(2)=21.13,df=12,p=0.0485),while no significant differences were found across hospital tiers(p>0.05).In terms of evaluation dimensions,accuracy was most frequently rated as“very important”(n=635).A significant association existed between hospital tier and the most valued dimension(χ^(2)=27.667,df=9,p=0.0011),with dermatologists in primary hospitals prioritizing traceability more than their peers in higher-tier hospitals.No significant associations were found across professional titles(p=0.127).Conclusions:Chinese dermatologists suggest a strong preference for ChatGPT-4o over domestic LLMs in psoriasis-related clinical tasks.While accuracy remains the primary criterion,traceability and logicality are also critical,particularly for clinicians in lower-tier hospitals.These findings suggest that future clinical LLMs should prioritize not only content accuracy but also source transparency and structural clarity to meet the diverse needs of different clinical settings.
基金supported by the National Natural Science Foundation of China(32471964)。
文摘The collection and annotation of lar ge-scale bird datasets are resource-intensive and time-consuming processes that significantly limit the scalability and accuracy of biodiversity monitoring systems.While self-supervised learning(SSL)has emerged as a promising approach for leveraging unannotated data,current SSL methods face two critical challenges in bird species recognition:(1)long-tailed data distributions that result in poor performance on underrepresented species;and(2)domain shift issues caused by data augmentation strategies designed to mitigate class imbalance.Here we present SDNet,a novel SSL-based bird recognition framework that integrates diffusion models with large language models(LLMs)to overcome these limitations.SDNet employs LLMs to generate semantically rich textual descriptions for tail-class species by prompting the models with species taxonomy,morphological attributes,and habitat information,producing detailed natural language priors that capture fine-grained visual characteristics(e.g.,plumage patterns,body proportions,and distinctive markings).These textual descriptions are subsequently used by a conditional diffusion model to synthesize new bird image samples through cross-attention mechanisms that fuse textual embeddings with intermediate visual feature representations during the denoising process,ensuring generated images preserve species-specific morphological details while maintaining photorealistic quality.Additionally,we incorporate a Swin Transformer as the feature extraction backbone whose hierarchical window-based attention mechanism and shifted windowing scheme enable multi-scale local feature extraction that proves particularly effective at capturing finegrained discriminative patterns(such as beak shape and feather texture)while mitigating domain shift between synthetic and original images through consistent feature representations across both data sources.SDNet is validated on both a self-constructed dataset(Bird_BXS)an d a publicly available benchmark(Birds_25),demonstrating substantial improvements over conventional SSL approaches.Our results indicate that the synergistic integration of LLMs,diffusion models,and the Swin Transformer architecture contributes significantly to recognition accuracy,particularly for rare and morphologically similar species.These findings highlight the potential of SDNet for addressing fundamental limitations of existing SSL methods in avian recognition tasks and establishing a new paradigm for efficient self-supervised learning in large-scale ornithological vision applications.
基金funded by the Office of the Vice-President for Research and Development of Cebu Technological University.
文摘This study demonstrates a novel integration of large language models,machine learning,and multicriteria decision-making to investigate self-moderation in small online communities,a topic under-explored compared to user behavior and platform-driven moderation on social media.The proposed methodological framework(1)utilizes large language models for social media post analysis and categorization,(2)employs k-means clustering for content characterization,and(3)incorporates the TODIM(Tomada de Decisão Interativa Multicritério)method to determine moderation strategies based on expert judgments.In general,the fully integrated framework leverages the strengths of these intelligent systems in a more systematic evaluation of large-scale decision problems.When applied in social media moderation,this approach promotes nuanced and context-sensitive self-moderation by taking into account factors such as cultural background and geographic location.The application of this framework is demonstrated within Facebook groups.Eight distinct content clusters encompassing safety,harassment,diversity,and misinformation are identified.Analysis revealed a preference for content removal across all clusters,suggesting a cautious approach towards potentially harmful content.However,the framework also highlights the use of other moderation actions,like account suspension,depending on the content category.These findings contribute to the growing body of research on self-moderation and offer valuable insights for creating safer and more inclusive online spaces within smaller communities.
基金National Key Research and Development Program of China(2024YFC3505400)Capital Clinical Project of Beijing Municipal Science&Technology Commission(Z221100007422092)Capital’s Funds for Health Improvement and Research(2024-1-2231).
文摘Objective To develop a clinical decision and prescription generation system(CDPGS)specifically for diarrhea in traditional Chinese medicine(TCM),utilizing a specialized large language model(LLM),Qwen-TCM-Dia,to standardize diagnostic processes and prescription generation.Methods Two primary datasets were constructed:an evaluation benchmark and a fine-tuning dataset consisting of fundamental diarrhea knowledge,medical records,and chain-ofthought(CoT)reasoning datasets.After an initial evaluation of 16 open-source LLMs across inference time,accuracy,and output quality,Qwen2.5 was selected as the base model due to its superior overall performance.We then employed a two-stage low-rank adaptation(LoRA)fine-tuning strategy,integrating continued pre-training on domain-specific knowledge with instruction fine-tuning using CoT-enriched medical records.This approach was designed to embed the clinical logic(symptoms→pathogenesis→therapeutic principles→prescriptions)into the model’s reasoning capabilities.The resulting fine-tuned model,specialized for TCM diarrhea,was designated as Qwen-TCM-Dia.Model performance was evaluated for disease diagnosis and syndrome type differentiation using accuracy,precision,recall,and F1-score.Furthermore,the quality of the generated prescriptions was compared with that of established open-source TCM LLMs.Results Qwen-TCM-Dia achieved peak performance compared to both the base Qwen2.5 model and five other open-source TCM LLMs.It achieved 97.05%accuracy and 91.48%F1-score in disease diagnosis,and 74.54%accuracy and 74.21%F1-score in syndrome type differentiation.Compared with existing open-source TCM LLMs(BianCang,HuangDi,LingDan,TCMLLM-PR,and ZhongJing),Qwen-TCM-Dia exhibited higher fidelity in reconstructing the“symptoms→pathogenesis→therapeutic principles→prescriptions”logic chain.It provided complete prescriptions,whereas other models often omitted dosages or generated mismatched prescriptions.Conclusion By integrating continued pre-training,CoT reasoning,and a two-stage fine-tuning strategy,this study establishes a CDPGS for diarrhea in TCM.The results demonstrate the synergistic effect of strengthening domain representation through pre-training and activating logical reasoning via CoT.This research not only provides critical technical support for the standardized diagnosis and treatment of diarrhea but also offers a scalable paradigm for the digital inheritance of expert TCM experience and the intelligent transformation of TCM.
基金supported in part by the National Natural Science Foundation of China under Grant 52432012in part by the Shanghai Science and Technology Project with 25ZR1402508。
文摘The use of Unmanned Aerial Vehicles(UAVs)for defect detection on railway slopes is becoming increasingly widespread due to their ability to capture high-resolution images over large,inaccessible,and topographically complex areas.However,current UAV-based detection methods face several critical limitations,including constrained deployment frequency,limited availability of annotated defect data,and the lack of mature risk assessment frameworks.To address these challenges,this study introduces a novel approach that integrates diffusion models with Large Language Models(LLMs)to generate highquality synthetic defect images tailored to railway slope scenarios.Furthermore,an improved transformerbased architecture is proposed,incorporating attention mechanisms and LLM-guided diffusion-generated imagery to enhance defect recognition performance under complex environmental conditions.Experimental evaluations conducted on a dataset of 300 field-collected images from high-risk railway slopes demonstrate that the proposed method significantly outperforms existing baselines in terms of precision,recall,and robustness,indicating strong applicability for real-world railway infrastructure monitoring and disaster prevention.
文摘Classifying job offers into occupational categories is a fundamental task in human resource information systems,as it improves and streamlines indexing,search,and matching between openings and job seekers.Comprehensive occupational databases such as O∗NET or ESCO provide detailed taxonomies of interrelated positions that can be leveraged to align the textual content of postings with occupational categories,thereby facilitating standardization,cross-system interoperability,and access to metadata for each occupation(e.g.,tasks,knowledge,skills,and abilities).In this work,we explore the effectiveness of fine-tuning existing language models(LMs)to classify job offers with occupational descriptors from O∗NET.This enables a more precise assessment of candidate suitability by identifying the specific knowledge and skills required for each position,and helps automate recruitment processes by mitigating human bias and subjectivity in candidate selection.We evaluate three representative BERT-like models:BERT,RoBERTa,and DeBERTa.BERT serves as the baseline encoder-only architecture;RoBERTa incorporates advances in pretraining objectives and data scale;and DeBERTa introduces architectural improvements through disentangled attention mechanisms.The best performance was achieved with the DeBERTa model,although the other models also produced strong results,and no statistically significant differences were observed acrossmodels.We also find that these models typically reach optimal performance after only a few training epochs,and that training with smaller,balanced datasets is effective.Consequently,comparable results can be obtained with models that require fewer computational resources and less training time,facilitating deployment and practical use.
文摘Knowledge distillation has become a standard technique for compressing large language models into efficient student models,but existing methods often struggle to balance prediction accuracy with explanation quality.Recent approaches such as Distilling Step-by-Step(DSbS)introduce explanation supervision,yet they apply it in a uniform manner that may not fully exploit the different learning dynamics of prediction and explanation.In this work,we propose a task-structured curriculum learning(TSCL)framework that structures training into three sequential phases:(i)prediction-only,to establish stable feature representations;(ii)joint prediction-explanation,to align task outputs with rationale generation;and(iii)explanation-only,to refine the quality of rationales.This design provides a simple but effective modification to DSbS,requiring no architectural changes and adding negligible training cost.We justify the phase scheduling with ablation studies and convergence analysis,showing that an initial prediction-heavy stage followed by a balanced joint phase improves both stability and explanation alignment.Extensive experiments on five datasets(e-SNLI,ANLI,CommonsenseQA,SVAMP,and MedNLI)demonstrate that TSCL consistently outperforms strong baselines,achieving gains of+1.7-2.6 points in accuracy and 0.8-1.2 in ROUGE-L,corresponding to relative error reductions of up to 21%.Beyond lexical metrics,human evaluation and ERASERstyle faithfulness diagnostics confirm that TSCL produces more faithful and informative explanations.Comparative training curves further reveal faster convergence and lower variance across seeds.Efficiency analysis shows less than 3%overhead in wall-clock training time and no additional inference cost,making the approach practical for realworld deployment.This study demonstrates that a simple task-structured curriculum can significantly improve the effectiveness of knowledge distillation.By separating and sequencing objectives,TSCL achieves a better balance between accuracy,stability,and explanation quality.The framework generalizes across domains,including medical NLI,and offers a principled recipe for future applications in multimodal reasoning and reinforcement learning.
基金the research project LaTe4PoliticES(PID2022-138099OB-I00)funded by MCIN/AEI/10.13039/501100011033 and the European Fund for Regional Development(ERDF)-a way to make Europe.Tomás Bernal-Beltrán is supported by University of Murcia through the predoctoral programme.
文摘The malicious dissemination of hate speech via compromised accounts,automated bot networks and malware-driven social media campaigns has become a growing cybersecurity concern.Automatically detecting such content in Spanish is challenging due to linguistic complexity and the scarcity of annotated resources.In this paper,we compare two predominant AI-based approaches for the forensic detection of malicious hate speech:(1)finetuning encoder-only models that have been trained in Spanish and(2)In-Context Learning techniques(Zero-and Few-Shot Learning)with large-scale language models.Our approach goes beyond binary classification,proposing a comprehensive,multidimensional evaluation that labels each text by:(1)type of speech,(2)recipient,(3)level of intensity(ordinal)and(4)targeted group(multi-label).Performance is evaluated using an annotated Spanish corpus,standard metrics such as precision,recall and F1-score and stability-oriented metrics to evaluate the stability of the transition from zero-shot to few-shot prompting(Zero-to-Few Shot Retention and Zero-to-Few Shot Gain)are applied.The results indicate that fine-tuned encoder-only models(notably MarIA and BETO variants)consistently deliver the strongest and most reliable performance:in our experiments their macro F1-scores lie roughly in the range of approximately 46%–66%depending on the task.Zero-shot approaches are much less stable and typically yield substantially lower performance(observed F1-scores range approximately 0%–39%),often producing invalid outputs in practice.Few-shot prompting(e.g.,Qwen 38B,Mistral 7B)generally improves stability and recall relative to pure zero-shot,bringing F1-scores into a moderate range of approximately 20%–51%but still falling short of fully fine-tuned models.These findings highlight the importance of supervised adaptation and discuss the potential of both paradigms as components in AI-powered cybersecurity and malware forensics systems designed to identify and mitigate coordinated online hate campaigns.
基金Guangzhou Science and Technology Program,Grant/Award Numbers:2025B03J0110,2024A03J1074,2024A03J0927。
文摘Large language models(LLMs)show considerable potential to revolutionize healthcare through their performance across diverse clinical applications.Given the inherent constraints of LLMs and the critical nature of medical practice,a rigorous and systematic evaluation of their medical competence is imperative.This study presents a comprehensive review of the established methodologies and benchmarks for evaluating the medical competence of LLMs,encompassing a thorough analysis of current assessment practices across medical knowledge,clinical practice competence,and ethical-safety considerations.By integrating clinician competency assessment frameworks into LLMs evaluation,we propose a structured tri-dimensional framework that systematically organizes existing evaluation approaches according to medical theoretical knowledge,clinical practice ability,and ethical-safety considerations.Furthermore,this research provides critical insights into future developmental trajectories while establishing foundational frameworks and standardization protocols for the integration of LLMs into medical practice.
