Rethinking Natural Language Generation with Layer-Wise Multi-View Decoding

Large Language Models (LLMs) have demonstrated significant potential in Medical Report Generation (MRG), yet their development requires large amounts of medical image-report pairs, which are commonly scattered across multiple centers. Centralizing these data is exceptionally challenging due to privacy regulations, thereby impeding model development and broader adoption of LLM-driven MRG models. To address this challenge, we present FedMRG, the first framework that leverages Federated Learning (FL) to enable privacy-preserving, multi-center development of LLM-driven MRG models, specifically designed to overcome the critical challenge of communication-efficient LLM training under multi-modal data heterogeneity. To start with, our framework tackles the fundamental challenge of communication overhead in federated LLM tuning by employing low-rank factorization to efficiently decompose parameter updates, significantly reducing gradient transmission costs and making LLM-driven MRG feasible in bandwidth-constrained FL settings. Furthermore, we observed the dual heterogeneity in MRG under the FL scenario: varying image characteristics across medical centers, as well as diverse reporting styles and terminology preferences. To address the data heterogeneity, we further enhance FedMRG with (1) client-aware contrastive learning in the MRG encoder, coupled with diagnosis-driven prompts, which capture both globally generalizable and locally distinctive features while maintaining diagnostic accuracy; and (2) a dual-adapter mutual boosting mechanism in the MRG decoder that harmonizes generic and specialized adapters to address variations in reporting styles and terminology. Through extensive evaluation of our established FL-MRG benchmark, we demonstrate the generalizability and adaptability of FedMRG, underscoring its potential in harnessing multi-center data and generating clinically accurate reports while maintaining communication efficiency.

Large language models (LLMs) are increasingly adopted in educational contexts to provide personalized support to students and teachers. The unprecedented capacity of LLM‐based applications to understand and generate natural language can potentially improve instructional effectiveness and learning outcomes, but the integration of LLMs in education technology has renewed concerns over algorithmic bias, which may exacerbate educational inequalities. Building on prior work that mapped the traditional machine learning life cycle, we provide a framework of the LLM life cycle from the initial development of LLMs to customizing pre‐trained models for various applications in educational settings. We explain each step in the LLM life cycle and identify potential sources of bias that may arise in the context of education. We discuss why current measures of bias from traditional machine learning fail to transfer to LLM‐generated text (eg, tutoring conversations) because text encodings are high‐dimensional, there can be multiple correct responses, and tailoring responses may be pedagogically desirable rather than unfair. The proposed framework clarifies the complex nature of bias in LLM applications and provides practical guidance for their evaluation to promote educational equity. Practitioner notes What is already known about this topic The life cycle of traditional machine learning (ML) applications which focus on predicting labels is well understood. Biases are known to enter in traditional ML applications at various points in the life cycle, and methods to measure and mitigate these biases have been developed and tested. Large language models (LLMs) and other forms of generative artificial intelligence (GenAI) are increasingly adopted in education technologies (EdTech), but current evaluation approaches are not specific to the domain of education. What this paper adds A holistic perspective of the LLM life cycle with domain‐specific examples in education to highlight opportunities and challenges for incorporating natural language understanding (NLU) and natural language generation (NLG) into EdTech. Potential sources of bias are identified in each step of the LLM life cycle and discussed in the context of education. A framework for understanding where to expect potential harms of LLMs for students, teachers, and other users of GenAI technology in education, which can guide approaches to bias measurement and mitigation. Implications for practice and/or policy Education practitioners and policymakers should be aware that biases can originate from a multitude of steps in the LLM life cycle, and the life cycle perspective offers them a heuristic for asking technology developers to explain each step to assess the risk of bias. Measuring the biases of systems that use LLMs in education is more complex than with traditional ML, in large part because the evaluation of natural language generation is highly context‐dependent (eg, what counts as good feedback on an assignment varies). EdTech developers can play an important role in collecting and curating datasets for the evaluation and benchmarking of LLM applications moving forward.

This article proposes a novel multitask learning framework that integrates contrastive learning and natural language generation (NLG) to enhance medical image classification and report generation. The goal is to improve disease classification accuracy and interpretability in medical diagnostics. The model architecture consists of a Vision Transformer (ViT) as a visual encoder, a transformer-based text encoder, and a multimodal decoder. The visual encoder processes medical images, while the text encoder handles disease-related text prompts. These components are trained jointly using image-text contrastive loss and language generation loss. Evaluations on the MIMICCXR and Chexpert datasets show that the model with NLG (Plain + NLG) outperforms the baseline contrastive learning model (Plain) in disease classification. For example, in the MIMICCXR dataset, the accuracy for Atelectasis increased from 17.44%(Plain) to 41.5% (Plain + NLG), and for Cardiomegaly, it improved from 19.25% to 47.4%. In Chexpert, the accuracy for Atelectasis increased from 12.5% to 58.5%, and for Pleural Effusion, from 61.10% to 64.0%. The model also demonstrated improvements in F1 scores, particularly for complex diseases like Cardiomegaly and Consolidation. The proposed multitask framework effectively combines contrastive learning with NLG, leading to improved disease classification and medical report generation. This approach has potential clinical applications by enhancing AI’s interpretability and accuracy in medical decision-making.

Large language models (LLMs) have shown impressive capabilities across various natural language tasks. However, evaluating their alignment with human preferences remains a challenge. To this end, we propose a comprehensive human evaluation framework to assess LLMs’ proficiency in following instructions on diverse real-world tasks. We construct a hierarchical task tree encompassing 7 major areas covering over 200 categories and over 800 tasks, which covers diverse capabilities such as question answering, reasoning, multiturn dialogue, and text generation, to evaluate LLMs in a comprehensive and in-depth manner. We also design detailed evaluation standards and processes to facilitate consistent, unbiased judgments from human evaluators. A test set of over 3,000 instances is released, spanning different difficulty levels and knowledge domains. Our work provides a standardized methodology to evaluate human alignment in LLMs for both English and Chinese. We also analyze the feasibility of automating parts of evaluation with a strong LLM (GPT-4). Our framework supports a thorough assessment of LLMs as they are integrated into real-world applications. We have made publicly available the task tree, TencentLLMEval dataset, and evaluation methodology which have been demonstrated as effective in assessing the performance of Tencent Hunyuan LLMs 1 . By doing so, we aim to facilitate the benchmarking of advances in the development of safe and human-aligned LLMs.

The article presents a method for identifying Russian-language texts generated by large language models (LLMs). The method was developed with a focus on short messages from 100 to 200 characters long. The relevance of the work is due to the widespread use of generative models, such as GPT-3.5, GPT-4o, LLaMA, GigaChat, DeepSeek, and Yandex GPT. The method is based on an ensemble of machine learning models, and features of three levels are also used: linguistic (structure, punctuation, morphology, lexical diversity), statistical (entropy, perplexity, n-gram frequency), and semantic (RuBERT embeddings). LightGBM, BiLSTM, and the pre-trained transformer model RuRoBERTa are used as basic models, combined by stacking through logistic regression. The choice of a hybrid ensemble approach is due to the desire to take into account features at different levels of the text hierarchy and to ensure the reliability of classification in the context of different topics of generated texts, versions, and types of language models. The use of an ensemble is an advantage in the analysis of short texts, since LightGBM, based on averaged indicators, is less sensitive to length (the perplexity metric is already averaged over the entire text), while BiLSTM and RoBERTa are able to identify local features of an LLM text, and not just global ones. The dataset of natural texts includes more than 2.8 million user comments from the VK social network. The LLM text dataset contains 700 thousand texts generated by seven relevant large language models. Topic modeling (LDA) and role generation using prompt engineering were used in the text generation. The methodology was evaluated on open datasets of Russian-language LLM texts. The experimental results showed an accuracy of up to 0.95 in the binary classification task (Human–LLM) and up to 0.89 in the multi-class task of determining the model-generator. The method demonstrates robustness to the diversity of sources, styles, and LLM versions.

Extracting information from unstructured and semi-structured geoscience literature is a crucial step in conducting geological research. The traditional machine learning extraction paradigm requires a substantial amount of high-quality manually annotated data for model training, which is time-consuming, labor-intensive, and not easily transferable to new fields. Recently, large language models (LLMs) (e.g., ChatGPT, GPT-4, and LLaMA), have shown great performance in various natural language processing (NLP) tasks, such as question answering, machine translation, and text generation. A substantial body of work has demonstrated that LLMs possess strong in-context learning (ICL) and even zero-shot learning capabilities to solve downstream tasks without specifically designed supervised fine-tuning. In this paper, we propose utilizing LLMs for geoscience literature information extraction. Specifically, we design a hierarchical PDF parsing pipeline and an automated knowledge extraction process, which can significantly reduce the need for manual data annotation, assisting geoscientists in literature data mining. For the hierarchical PDF parsing pipeline, firstly, a document layout detection model fine-tuned on geoscience literature is employed for layout detection, obtaining layout detection information for the document. Secondly, based on the document layout information, an optical character content parsing model is used for content parsing, obtaining the text structure and plain text corresponding to the content. Finally, the text structure and plain text are combined and reconstructed to ultimately obtain the parsed structured data. For the automated knowledge extraction process, firstly, the parsed long text is segmented into paragraphs to adapt to the input length limit of LLMs. Subsequently, a few-shot prompting method is employed for structured knowledge extraction, encompassing two tasks: attribute value extraction and triplet extraction. In attribute value extraction, prompts are generated automatically by the LLMs based on the subdomain and attribute names, facilitating the location and extraction of values related to subdomain attribute names in the text. For triplet extraction, the LLMs employ a procedural approach to entity extraction, entity type extraction, and relation extraction, following the knowledge graph structure pattern. Finally, the extracted structured knowledge is stored in the form of knowledge graphs, facilitating further analysis and integration of various types of knowledge from the literature. Our proposed approach turns out to be simple, flexible, and highly effective in geoscience literature information extraction. Demonstrations of information extraction in subdomains such as radiolarian fossils and fluvial facies have yielded satisfactory results. The extraction efficiency has significantly improved, and feedback from domain experts indicates a relatively high level of accuracy in the extraction process. The extracted results can be used to construct a foundational knowledge graph for geoscience literature, supporting the comprehensive construction and efficient application of a geoscience knowledge graph.

Large language models (LLMs) have emerged as a milestone in the realm of machine learning, offering groundbreaking capabilities in understanding human language with remarkable coherence and versatility. Their application spans a multitude of domains, ranging from text classification to information extraction. Such models are first pre-trained on vast data corpora to learn general language patterns and then are fine-tuned on specific downstream tasks. Despite their impressive capabilities, LLMs encounter several limitations, including sub-optimal performance in understanding the domain-specific language of drug labels, limited performance in analyzing class-imbalanced data, and high computational complexity of training. Human prescription drug labeling contains a summary of the essential scientific information needed for the safe and effective use of the drug and includes critical information about drug products, such as pharmacokinetics and adverse events. Automatic information extraction from drug labels can facilitate processes such as finding the adverse reaction of the drugs or finding the interaction of one drug with another drug. Although LLMs are an interesting solution to perform this task, because of their training on generic data, they may not possess the depth of knowledge or the ability required to understand the language in such a specialized field. To address this issue, this work presents the developed PharmBERT, which is a BERT model specifically pretrained on the drug labels, publicly available online for the NLP community. We demonstrate that our model outperforms models such as vanilla BERT, ClinicalBERT, and BioBERT in multiple NLP tasks in the drug label domain. Another shortcoming of LLMs lies in their handling of class-imbalanced data, leading to a subpar performance in identifying and accurately responding to minority classes. This class-imbalance can skew the model's learning process, favoring majority classes, and thus diminishing its effectiveness in recognizing and appropriately addressing less represented categories. To cope with these challenges, we present a simple modification of standard fine-tuning. Specifically, we propose a two-stage fine-tuning. In Stage 1, we fine-tune the final layer of the pretrained model with class-balanced augmented data, generated using ChatGPT. As a large generative language model, ChatGPT is capable of generating novel yet contextually similar responses to a given text sample, which makes it an excellent candidate for data augmentation. In Stage 2, we perform the standard fine-tuning. Our modification has several benefits: (1) it leverages pretrained representations by only fine-tuning a small portion of the model parameters while keeping the rest untouched; (2) it allows the model to learn an initial representation of the specific task; and importantly (3) it protects the learning of tail classes from being at a disadvantage during the model updating. The experimental results show that the proposed two-stage fine-tuning outperforms vanilla fine-tuning and state-of-the-art methods on the above datasets. The third issue of LLMs is the high computational complexity of training. Due to the large number of parameters in many state-of-the-art LLMs, such as Bidirectional Encoder Representations from Transformers (BERT), the process of fine-tuning exhibits a high level of intricacy. One attractive solution to this issue is parameter-efficient fine-tuning, which means only fine-tuning a small portion of the model and not changing the rest. However, the question that which portion of the BERT model is the most important one for fine-tuning remains unanswered. Here, we first analyze different components in the BERT model to see which one undergoes the most change after fine-tuning. We show that output LayerNorm changes more than other components when finetuned for different General Language Understanding Evaluation (GLUE) tasks. Then we show that only fine-tuning this component reaches a comparable, or in some cases better, performance to full fine-tuning and other parameter-efficient fine-tuning methods. Moreover, we use Fisher information to find out which parameters in LayerNorm are the most important ones and show that with a small performance degradation, many NLP tasks in the GLUE benchmark can be solved by fine-tuning only a small, selected portion of LayerNorm. In summary, LLMs hold immense potential to redefine and enhance various fields, including drug-label analysis, but they also suffer from limitations like sub-optimal performance in understanding domain-specific texts of drug labels, not learning minority classes in the class-imbalanced data, and high computational complexity. In this thesis, for each one of these shortcomings, first, the problem is analyzed and then a solution is provided.

Purpose This study aims to explore the integration of large language models (LLMs) and vision-language models (VLMs) in robotics, highlighting their potential benefits and the safety challenges they introduce, including robustness issues, adversarial vulnerabilities, privacy concerns and ethical implications. Design/methodology/approach This survey conducts a comprehensive analysis of the safety risks associated with LLM- and VLM-powered robotic systems. The authors review existing literature, analyze key challenges, evaluate current mitigation strategies and propose future research directions. Findings The study identifies that ensuring the safety of LLM-/VLM-driven robots requires a multi-faceted approach. While current mitigation strategies address certain risks, gaps remain in real-time monitoring, adversarial robustness and ethical safeguards. Originality/value This study offers a structured and comprehensive overview of the safety challenges in LLM-/VLM-driven robotics. It contributes to ongoing discussions by integrating technical, ethical and regulatory perspectives to guide future advancements in safe and responsible artificial intelligence-driven robotics.

The development of large language models (LLMs) such as ChatGPT has brought a lot of attention recently. However, their evaluation in the benchmark academic datasets remains under-explored due to the difficulty of evaluating the generative outputs produced by this model against the ground truth. In this paper, we aim to present a thorough evaluation of ChatGPT's performance on diverse academic datasets, covering tasks like question-answering, text summarization, code generation, commonsense reasoning, mathematical problem-solving, machine translation, bias detection, and ethical considerations. Specifically, we evaluate ChatGPT across 140 tasks and analyze 255K responses it generates in these datasets. This makes our work the largest evaluation of ChatGPT in NLP benchmarks. In short, our study aims to validate the strengths and weaknesses of ChatGPT in various tasks and provide insights for future research using LLMs. We also report a new emergent ability to follow multi-query instructions that we mostly found in ChatGPT and other instruction-tuned models. Our extensive evaluation shows that even though ChatGPT is capable of performing a wide variety of tasks, and may obtain impressive performance in several benchmark datasets, it is still far from achieving the ability to reliably solve many challenging tasks. By providing a thorough assessment of ChatGPT's performance across diverse NLP tasks, this paper sets the stage for a targeted deployment of ChatGPT-like LLMs in real-world applications.

Biases in language and their reflection in language models have attracted researchers' attention, particularly with the growth of large language models (LLMs). However, many questions on the links between language models and people’s biased attitudes remain unanswered. In the current study we focus on gender–career bias to examine the extent to which language models can be used to model behavioural responses in the Gender–Career Implicit Association Test (IAT). We provide a systematic evaluation of a range of language models, including n-gram, count vector, predict (word2vec), and Large Language Models (LLMs), to determine how well they capture people’s behaviour in the IAT. We compared response time data from over 800,000 participants against 25 language models, with a total of 675 model variants. We find that many language models, including large language models (LLMs), correlated well with human behavior. While results support previous findings for both predict and count model families, we observed that performance of LLMs was consistently different from that of simpler predict models, particularly in terms of the direction and strength of correlations with reaction time and bias. This divergence may indicate successful attempts to mitigate bias in LLMs while preserving other aspects of linguistic information. Our findings reinforce the idea that societal biases are generally encoded in language, but that large language models can exhibit behaviors different to classical language models.

Large language models have demonstrated remarkable capabilities in facilitating general human-computer interactions. However, their effectiveness in providing domain-specific information in professional fields is limited. In this paper, we present an intelligent assisted decision-making framework utilizing large-language models to provide professional advice in taxation. By augmenting the large language model with a domain-specific database, we aim to overcome the limitations of generic language models and provide professional advice for tax-related queries. The results of our study indicate that the proposed framework sur-passes the credibility of generic large language models by providing domain-specific advice in taxation.

Abstract. Large language models (LLMs) have revolutionized the field of natural language processing (NLP), demonstrating remarkable capabilities in understanding, generating, and manipulating human language. This comprehensive review explores the development, applications, optimizations, and challenges of LLMs. This paper begin by tracing the evolution of these models and their foundational architectures, such as the Transformer, GPT, and BERT. We then delve into the applications of LLMs in natural language understanding tasks, including sentiment analysis, named entity recognition, question answering, and text summarization, highlighting real-world use cases. Next, we examine the role of LLMs in natural language generation, covering areas such as content creation, language translation, personalized recommendations, and automated responses. We further discuss LLM applications in other NLP tasks like text style transfer, text correction, and language model pre-training. Subsequently, we explore techniques for optimizing and improving LLMs, including model compression, explainability, robustness, and security. Finally, we address the challenges posed by the significant computational requirements, sample inefficiency, and ethical considerations surrounding LLMs. We conclude by discussing potential future research directions, such as efficient architectures, few-shot learning, bias mitigation, and privacy-preserving techniques, which will shape the ongoing development and responsible deployment of LLMs in NLP.

With large-scale language models demonstrating superior capabilities in a wide range of downstream natural language processing tasks, the future trajectory of research in the field of text categorization faces increasing uncertainty. In this evolving paradigm of open-ended language modeling, where task delimitations are increasingly blurred, a pressing question arises: to what extent has text classification advanced under the full potential of large language model (LLM)? To address this pivotal inquiry, we introduce recurrent generative pre-trained transformer (RGPT), an adaptive boosting framework meticulously designed to craft a dedicated LLM for text classification. RGPT constructs a sequence of base learners by dynamically modulating the training data distribution and iteratively fine-tuning LLMs. These base learners are then progressively integrated, leveraging historical prediction trajectories to form a highly specialized text classification model. Extensive empirical evaluations demonstrate that RGPT surpasses eight state-of-the-art pretrained language models and seven cutting-edge LLMs across four benchmark datasets, achieving an average performance gain of 2.90%.

The subject of the study is the analysis and improvement of methods for determining the relevance of project names to the information content of purchases using large language models. The object of the study is a database containing the names of projects and purchases in the field of electric power industry, collected from open sources. The author examines in detail such aspects of the topic as the use of TF-IDF and cosine similarity metrics for primary data filtering, and also describes in detail the integration and evaluation of the effectiveness of large language models such as GigaChat, GPT-3.5, and GPT-4 in text data matching tasks. Special attention is paid to the methods of clarifying the similarity of names based on reflection introduced into the prompta of large language models, which makes it possible to increase the accuracy of data comparison. The study uses TF-IDF and cosine similarity methods for primary data analysis, as well as large GigaChat, GPT-3.5 and GPT-4 language models for detailed verification of the relevance of project names and purchases, including reflection in model prompta to improve the accuracy of results. The novelty of the research lies in the development of a combined approach to determining the relevance of project names and purchases, combining traditional methods of processing text information (TF-IDF, cosine similarity) with the capabilities of large language models. A special contribution of the author to the research of the topic is the proposed methodology for improving the accuracy of data comparison by clarifying the results of primary selection using GPT-3.5 and GPT-4 models with optimized prompta, including reflection. The main conclusions of the study are confirmation of the prospects of using the developed approach in the tasks of information support for procurement processes and project implementation, as well as the possibility of using the results obtained for the development of text data mining systems in various sectors of the economy. The study showed that the use of language models makes it possible to improve the value of the F2 measure to 0.65, which indicates a significant improvement in the quality of data comparison compared with basic methods.

Large Language Models (LLMs) have garnered increasing attention in neurosurgery and possess significant potential to improve the field. However, the breadth and performance of LLMs across diverse neurosurgical tasks have not been systematically examined, and LLMs come with their own challenges and unique terminology. We seek to identify key models, establish reporting guidelines for replicability, and highlight progress in key application areas of LLM use in the neurosurgical literature. We searched PubMed and Google Scholar using terms related to LLMs and neurosurgery ("large language model" OR "LLM" OR "ChatGPT" OR "GPT-3" OR "GPT3" OR "GPT-3.5" OR "GPT3.5" OR "GPT-4" OR "GPT4" OR "LLAMA" OR "MISTRAL" OR "BARD") AND "neurosurgery". The final set of articles was reviewed for publication year, application area, specific LLM(s) used, control/comparison groups used to evaluate LLM performance, whether the article reported specific LLM prompts, prompting strategy types used, whether the LLM query could be reproduced in its entirety (including both the prompt used and any adjoining data), measures of hallucination, and reported performance measures. Fifty-one articles met inclusion criteria, and were categorized into six application areas, with the most common being Generation of Text for Direct Clinical Use (n = 14, 27.5%), Answering Standardized Exam Questions (n = 12, 23.5%), and Clinical Judgement and Decision-Making Support (n = 11, 21.6%). The most frequently used LLMs were GPT-3.5 (n = 30, 58.8%), GPT-4 (n = 20, 39.2%), Bard (n = 9, 17.6%), and Bing (n = 6, 11.8%). Most studies (n = 43, 84.3%) used LLMs directly out-of-the-box, while 8 studies (15.7%) conducted advanced pre-training or fine-tuning. Large language models show advanced capabilities in complex tasks and hold potential to transform neurosurgery. However, research typically addresses basic applications and overlooks enhancing LLM performance, facing reproducibility issues. Standardizing detailed reporting, considering LLM stochasticity, and using advanced methods beyond basic validation are essential for progress.