The astonishing performance of large language models (LLMs) and their remarkable achievements in production and daily life have led to their widespread application in collaborative tasks. However, current large models face challenges such as hallucination and lack of specificity in content generation in vertical domain tasks. Inspired by the contrast and classification mechanisms in human cognitive processes, this paper constructs an adversarial learning-based prompt framework named ChallengeMe, which includes three cascaded solutions: generation prompts, evaluation prompts, and feedback optimization. In this process, we designed seven core optimization dimensions and set the threshold for adversarial learning. The results of mixed case studies on the text summarization task show that the proposed framework can generate more accurate and fluent text summaries compared to the current advanced mainstream LLMs.
Large language models (LLMs) have facilitated the generation of high-quality, cost-effective synthetic data for developing downstream models and conducting statistical analyses in various domains. However, the increased reliance on synthetic data may pose potential negative impacts. Numerous studies have demonstrated that LLM-generated synthetic data can perpetuate and even amplify societal biases and stereotypes, and produce erroneous outputs known as ``hallucinations'' that deviate from factual knowledge. In this paper, we aim to audit artifacts, such as classifiers, generators, or statistical plots, to identify those trained on or derived from synthetic data and raise user awareness, thereby reducing unexpected consequences and risks in downstream applications. To this end, we take the first step to introduce synthetic artifact auditing to assess whether a given artifact is derived from LLM-generated synthetic data. We then propose an auditing framework with three methods including metric-based auditing, tuning-based auditing, and classification-based auditing. These methods operate without requiring the artifact owner to disclose proprietary training details. We evaluate our auditing framework on three text classification tasks, two text summarization tasks, and two data visualization tasks across three training scenarios. Our evaluation demonstrates the effectiveness of all proposed auditing methods across all these tasks. For instance, black-box metric-based auditing can achieve an average accuracy of $0.868 \pm 0.071$ for auditing classifiers and $0.880 \pm 0.052$ for auditing generators using only 200 random queries across three scenarios. We hope our research will enhance model transparency and regulatory compliance, ensuring the ethical and responsible use of synthetic data.
Recent advancements in large language models (LLMs) have shown potential for transforming data processing in healthcare, particularly in understanding complex clinical narratives. This study evaluates the efficacy of zero-shot LLMs in summarizing long clinical texts that require temporal reasoning, a critical aspect for comprehensively capturing patient histories and treatment trajectories. We applied a series of advanced zero-shot LLMs to extensive clinical documents, assessing their ability to integrate and accurately reflect temporal dynamics without prior task-specific training. While the models efficiently identified key temporal events, they struggled with chronological coherence over prolonged narratives. The evaluation, combining quantitative and qualitative methods, highlights the strengths and limitations of zero-shot LLMs in clinical text summarization. The results suggest that while promising, zero-shot LLMs require further refinement to effectively support clinical decision-making processes, underscoring the need for enhanced model training approaches that better capture the nuances of temporal information in long context medical documents.
Given the recent introduction of multiple language models and the ongoing demand for improved Natural Language Processing tasks, particularly summarization, this work provides a comprehensive benchmarking of 20 recent language models, focusing on smaller ones for the news summarization task. In this work, we systematically test the capabilities and effectiveness of these models in summarizing news article texts which are written in different styles and presented in three distinct datasets. Specifically, we focus in this study on zero-shot and few-shot learning settings and we apply a robust evaluation methodology that combines different evaluation concepts including automatic metrics, human evaluation, and LLM-as-a-judge. Interestingly, including demonstration examples in the few-shot learning setting did not enhance models' performance and, in some cases, even led to worse quality of the generated summaries. This issue arises mainly due to the poor quality of the gold summaries that have been used as reference summaries, which negatively impacts the models' performance. Furthermore, our study's results highlight the exceptional performance of GPT-3.5-Turbo and GPT-4, which generally dominate due to their advanced capabilities. However, among the public models evaluated, certain models such as Qwen1.5-7B, SOLAR-10.7B-Instruct-v1.0, Meta-Llama-3-8B and Zephyr-7B-Beta demonstrated promising results. These models showed significant potential, positioning them as competitive alternatives to large models for the task of news summarization.
Artificial intelligence systems significantly impact the environment, particularly in natural language processing (NLP) tasks. These tasks often require extensive computational resources to train deep neural networks, including large-scale language models containing billions of parameters. This study analyzes the trade-offs between energy consumption and performance across three neural language models: two pre-trained models (T5-base and BART-base), and one large language model (LLaMA 3-8B). These models were fine-tuned for the text summarization task, focusing on generating research paper highlights that encapsulate the core themes of each paper. A wide range of evaluation metrics, including ROUGE, METEOR, MoverScore, BERTScore, and SciBERTScore, were employed to assess their performance. Furthermore, the carbon footprint associated with fine-tuning each model was measured, offering a comprehensive assessment of their environmental impact. This research underscores the importance of incorporating environmental considerations into the design and implementation of neural language models and calls for the advancement of energy-efficient AI methodologies.
Text summarization involves reducing extensive documents to short sentences that encapsulate the essential ideas. The goal is to create a summary that effectively conveys the main points of the original text. We spend a significant amount of time each day reading the newspaper to stay informed about current events both domestically and internationally. While reading newspapers enriches our knowledge, we sometimes come across unnecessary content that isn't particularly relevant to our lives. In this paper, we introduce a neural network model designed to summarize Bangla text into concise and straightforward paragraphs, aiming for greater stability and efficiency.
This study explores the overlap between text summarization and simplification outputs. While summarization evaluation methods are streamlined, simplification lacks cohesion, prompting the question: how closely can abstractive summarization resemble gold-standard simplification? We address this by applying two BART-based BRIO summarization methods to the Newsela corpus, comparing outputs with manually annotated simplifications and achieving a top ROUGE-L score of 0.654. This provides insight into where summarization and simplification outputs converge and differ.
Recently, multimodal depression recognition for clinical interviews (MDRC) has recently attracted considerable attention. Existing MDRC studies mainly focus on improving task performance and have achieved significant development. However, for clinical applications, model transparency is critical, and previous works ignore the interpretability of decision-making processes. To address this issue, we propose an Explainable Multimodal Depression Recognition for Clinical Interviews (EMDRC) task, which aims to provide evidence for depression recognition by summarizing symptoms and uncovering underlying causes. Given an interviewer-participant interaction scenario, the goal of EMDRC is to structured summarize participant's symptoms based on the eight-item Patient Health Questionnaire depression scale (PHQ-8), and predict their depression severity. To tackle the EMDRC task, we construct a new dataset based on an existing MDRC dataset. Moreover, we utilize the PHQ-8 and propose a PHQ-aware multimodal multi-task learning framework, which captures the utterance-level symptom-related semantic information to help generate dialogue-level summary. Experiment results on our annotated dataset demonstrate the superiority of our proposed methods over baseline systems on the EMDRC task.
This paper introduces an information-theoretic framework for text summarization. We define the summarizer rate-distortion function and show that it provides a fundamental lower bound on summarizer performance. We describe an iterative procedure, similar to Blahut-Arimoto algorithm, for computing this function. To handle real-world text datasets, we also propose a practical method that can calculate the summarizer rate-distortion function with limited data. Finally, we empirically confirm our theoretical results by comparing the summarizer rate-distortion function with the performances of different summarizers used in practice.
Reinforcement learning from human feedback (RLHF) has proven effective in aligning large language models (LLMs) with human preferences, but often at the cost of reduced output diversity. This trade-off between diversity and alignment quality remains a significant challenge. Drawing inspiration from curiosity-driven exploration in reinforcement learning, we introduce curiosity-driven RLHF (CD-RLHF), a framework that incorporates intrinsic rewards for novel states, alongside traditional sparse extrinsic rewards, to optimize both output diversity and alignment quality. We demonstrate the effectiveness of CD-RLHF through extensive experiments on a range of tasks, including text summarization and instruction following. Our approach achieves significant gains in diversity on multiple diversity-oriented metrics while maintaining alignment with human preferences comparable to standard RLHF. We make our code publicly available at https://github.com/ernie-research/CD-RLHF.
The field of Artificial Intelligence (AI) continues to drive transformative innovations, with significant progress in conversational interfaces, autonomous vehicles, and intelligent content creation. Since the launch of ChatGPT in late 2022, the rise of Generative AI has marked a pivotal era, with the term Large Language Models (LLMs) becoming a ubiquitous part of daily life. LLMs have demonstrated exceptional capabilities in tasks such as text summarization, code generation, and creative writing. However, these models are inherently limited by their token-level processing, which restricts their ability to perform abstract reasoning, conceptual understanding, and efficient generation of long-form content. To address these limitations, Meta has introduced Large Concept Models (LCMs), representing a significant shift from traditional token-based frameworks. LCMs use concepts as foundational units of understanding, enabling more sophisticated semantic reasoning and context-aware decision-making. Given the limited academic research on this emerging technology, our study aims to bridge the knowledge gap by collecting, analyzing, and synthesizing existing grey literature to provide a comprehensive understanding of LCMs. Specifically, we (i) identify and describe the features that distinguish LCMs from LLMs, (ii) explore potential applications of LCMs across multiple domains, and (iii) propose future research directions and practical strategies to advance LCM development and adoption.
Aligning language models (LMs) with preferences is an important problem in natural language generation. A key challenge is that preferences are typically provided at the sequence level while LM training and generation both occur at the token level. There is, therefore, a granularity mismatch between the preference and the LM training losses, which may complicate the learning problem. In this paper, we address this issue by developing an alternate training process, where we iterate between grounding the sequence-level preference into token-level training guidance, and improving the LM with the learned guidance. For guidance learning, we design a framework that extends the pairwise-preference learning in imitation learning to both variable-length LM generation and the utilization of the preference among multiple generations. For LM training, based on the amount of supervised data, we present two minimalist learning objectives that utilize the learned guidance. In experiments, our method performs competitively on two distinct representative LM tasks -- discrete-prompt generation and text summarization.
Scientific figure captioning is a complex task that requires generating contextually appropriate descriptions of visual content. However, existing methods often fall short by utilizing incomplete information, treating the task solely as either an image-to-text or text summarization problem. This limitation hinders the generation of high-quality captions that fully capture the necessary details. Moreover, existing data sourced from arXiv papers contain low-quality captions, posing significant challenges for training large language models (LLMs). In this paper, we introduce a framework called Multi-LLM Collaborative Figure Caption Generation (MLBCAP) to address these challenges by leveraging specialized LLMs for distinct sub-tasks. Our approach unfolds in three key modules: (Quality Assessment) We utilize multimodal LLMs to assess the quality of training data, enabling the filtration of low-quality captions. (Diverse Caption Generation) We then employ a strategy of fine-tuning/prompting multiple LLMs on the captioning task to generate candidate captions. (Judgment) Lastly, we prompt a prominent LLM to select the highest quality caption from the candidates, followed by refining any remaining inaccuracies. Human evaluations demonstrate that informative captions produced by our approach rank better than human-written captions, highlighting its effectiveness. Our code is available at https://github.com/teamreboott/MLBCAP
This thesis presents Abstractive Text Summarization models for contemporary Sanskrit prose. The first chapter, titled Introduction, presents the motivation behind this work, the research questions, and the conceptual framework. Sanskrit is a low-resource inflectional language. The key research question that this thesis investigates is what the challenges in developing an abstractive TS for Sanskrit. To answer the key research questions, sub-questions based on four different themes have been posed in this work. The second chapter, Literature Review, surveys the previous works done. The third chapter, data preparation, answers the remaining three questions from the third theme. It reports the data collection and preprocessing challenges for both language model and summarization model trainings. The fourth chapter reports the training and inference of models and the results obtained therein. This research has initiated a pipeline for Sanskrit abstractive text summarization and has reported the challenges faced at every stage of the development. The research questions based on every theme have been answered to answer the key research question.
Automatic text summarization, particularly headline generation, remains a critical yet underexplored area for Bengali religious news. Existing approaches to headline generation typically rely solely on the article content, overlooking crucial contextual features such as sentiment, category, and aspect. This limitation significantly hinders their effectiveness and overall performance. This study addresses this limitation by introducing a novel corpus, BeliN (Bengali Religious News) - comprising religious news articles from prominent Bangladeshi online newspapers, and MultiGen - a contextual multi-input feature fusion headline generation approach. Leveraging transformer-based pre-trained language models such as BanglaT5, mBART, mT5, and mT0, MultiGen integrates additional contextual features - including category, aspect, and sentiment - with the news content. This fusion enables the model to capture critical contextual information often overlooked by traditional methods. Experimental results demonstrate the superiority of MultiGen over the baseline approach that uses only news content, achieving a BLEU score of 18.61 and ROUGE-L score of 24.19, compared to baseline approach scores of 16.08 and 23.08, respectively. These findings underscore the importance of incorporating contextual features in headline generation for low-resource languages. By bridging linguistic and cultural gaps, this research advances natural language processing for Bengali and other underrepresented languages. To promote reproducibility and further exploration, the dataset and implementation code are publicly accessible at https://github.com/akabircs/BeliN.
India's vast linguistic diversity presents unique challenges and opportunities for technological advancement, especially in the realm of Natural Language Processing (NLP). While there has been significant progress in NLP applications for widely spoken languages, the regional languages of India, such as Marathi and Hindi, remain underserved. Research in the field of NLP for Indian regional languages is at a formative stage and holds immense significance. The paper aims to build a platform which enables the user to use various features like text anonymization, abstractive text summarization and spell checking in English, Hindi and Marathi language. The aim of these tools is to serve enterprise and consumer clients who predominantly use Indian Regional Languages.
The advancements in deep learning, particularly the introduction of transformers, have been pivotal in enhancing various natural language processing (NLP) tasks. These include text-to-text applications such as machine translation, text classification, and text summarization, as well as data-to-text tasks like response generation and image-to-text tasks such as captioning. Transformer models are distinguished by their attention mechanisms, pretraining on general knowledge, and fine-tuning for downstream tasks. This has led to significant improvements, particularly in abstractive summarization, where sections of a source document are paraphrased to produce summaries that closely resemble human expression. The effectiveness of these models is assessed using diverse metrics, encompassing techniques like semantic overlap and factual correctness. This survey examines the state of the art in text summarization models, with a specific focus on the abstractive summarization approach. It reviews various datasets and evaluation metrics used to measure model performance. Additionally, it includes the results of test cases using abstractive summarization models to underscore the advantages and limitations of contemporary transformer-based models. The source codes and the data are available at https://github.com/gospelnnadi/Text-Summarization-SOTA-Experiment.
In this work, we propose a Multi-LLM summarization framework, and investigate two different multi-LLM strategies including centralized and decentralized. Our multi-LLM summarization framework has two fundamentally important steps at each round of conversation: generation and evaluation. These steps are different depending on whether our multi-LLM decentralized summarization is used or centralized. In both our multi-LLM decentralized and centralized strategies, we have k different LLMs that generate diverse summaries of the text. However, during evaluation, our multi-LLM centralized summarization approach leverages a single LLM to evaluate the summaries and select the best one whereas k LLMs are used for decentralized multi-LLM summarization. Overall, we find that our multi-LLM summarization approaches significantly outperform the baselines that leverage only a single LLM by up to 3x. These results indicate the effectiveness of multi-LLM approaches for summarization.
The NAHU$^2$ project is a Franco-Mexican collaboration aimed at building the $\pi$-YALLI corpus adapted to machine learning, which will subsequently be used to develop computer resources for the Nahuatl language. Nahuatl is a language with few computational resources, even though it is a living language spoken by around 2 million people. We have decided to build $\pi$-YALLI, a corpus that will enable to carry out research on Nahuatl in order to develop Language Models (LM), whether dynamic or not, which will make it possible to in turn enable the development of Natural Language Processing (NLP) tools such as: a) a grapheme unifier, b) a word segmenter, c) a POS grammatical analyser, d) a content-based Automatic Text Summarization; and possibly, e) a translator translator (probabilistic or learning-based).
Text-to-image models are vulnerable to the stepwise "Divide-and-Conquer Attack" (DACA) that utilize a large language model to obfuscate inappropriate content in prompts by wrapping sensitive text in a benign narrative. To mitigate stepwise DACA attacks, we propose a two-layer method involving text summarization followed by binary classification. We assembled the Adversarial Text-to-Image Prompt (ATTIP) dataset ($N=940$), which contained DACA-obfuscated and non-obfuscated prompts. From the ATTIP dataset, we created two summarized versions: one generated by a small encoder model and the other by a large language model. Then, we used an encoder classifier and a GPT-4o classifier to perform content moderation on the summarized and unsummarized prompts. When compared with a classifier that operated over the unsummarized data, our method improved F1 score performance by 31%. Further, the highest recorded F1 score achieved (98%) was produced by the encoder classifier on a summarized ATTIP variant. This study indicates that pre-classification text summarization can inoculate content detection models against stepwise DACA obfuscations.