Language Understanding Research Articles

Examining Stakeholders’ Prospects on the Implementation of Equity-Based Strategies in Public Secondary Schools in Tanzania

The desire to address equity issues in the education sector is a shared philosophy by a number of countries, including Tanzania. As a result, Tanzania as a country is striving to address the issue of equity in its education system through the implementation of various equity strategies. This paper therefore sought to examine stakeholders’ prospects on the implementation of equity-based strategies in public secondary schools in Tanzania. Precisely, the interest of the researchers was to explore what was being done in the selected public secondary schools to implement equity-based strategies. A mixed-methods research approach combining both qualitative and quantitative analysis approaches was employed. The study also used a multiple case study design, which allowed the selection of two case study areas i.e. Dodoma City representing the urban setting, and Kondoa district representing the rural setting. A total of One Hundred Forty-Four (144) participants from seven (7) public secondary schools were involved in the study. Data were collected through interviews and questionnaires. Quantitative data were analyzed through descriptive analysis and presented in numerical and graphical form while qualitative data were analyzed through inductive content analysis and presented in thematic form and narratives. Study findings revealed that the selected study schools were indeed implementing a number of activities towards the realization of equity and thus the prospects for implementing equity based strategies in these schools was found to be positive. For instance, activities such as the use of common and understandable language i.e., English and Kiswahili; ensuring that schools have qualified teaching staff, ensuring a supportive learning environment as well as strengthening and supporting school leadership were being implemented. Nevertheless, the study also noted that the implementation of some key factors critical for realization of equity in the schools such as incentive provision to teachers, use of appropriate technology and use of different learning styles was still weak and thus needed improvement.

Reconstructing representations using diffusion models for multimodal sentiment analysis through reading comprehension

The primary challenge in multimodal sentiment analysis (MSA), which utilizes textual, audio, and visual information to analyze speakers' emotions, lies in constructing representation vectors that incorporate both unimodal semantic and multimodal interaction information. While existing research has extensively focused on multimodal fusion strategies, there remains insufficient exploration of the intrinsic potential within concurrently enhancing unimodal and multimodal representations. To address this gap, we introduce two additional steps to traditional three-step MSA: text modality enhancement through the machine reading comprehension (MRC) framework and multimodal representation reconstruction via proposing the diverse diffusion denoising autoencoder (D3AE) module. The MRC queries are integrated to locate sentiment-related prior knowledge, thereby deepening textual semantic understanding from a pretrained language model. Meanwhile, D3AE employs a single-step denoising strategy along with diffusion models across multiple time intervals, enabling efficient reconstruction and enhancement of multimodal representations. Extensive experiments conducted on two benchmark datasets, CMU-MOSI and CMU-MOSEI, validate that our model, MRC-D3AE, achieves state-of-the-art performance. The superiority of our model over existing baselines is primarily attributed to integrating MRC for enhancing text modality and D3AE for reconstructing multimodal representations.

How Does Understanding of Social Situations and Other's Intention Contribute to Idiom and Irony Comprehension in Autism Spectrum Disorder?

Background/Objectives: Figurative language is a central tool for enriching spoken and written languages, and it is important for building social relationships. Difficulties in figurative language understanding may impair social adjustment. Some studies have found more gaps in the understanding of irony and idioms among children with autism spectrum disorder (ASD) compared to those of their peers with typical development (TD). To date, no studies have examined the relationship between the ability to understand social situations (as a separate ability) and the ability to understand irony and idioms. The present study examined the roles of theory of mind (ToM) and social situation understanding in the comprehension of idioms and ironic criticism. Methods: The current study included 58 participants aged 8-11, including 28 children with high-functioning ASD and 30 children with TD matched by age, gender, and nonverbal intelligence. All the participants completed a ToM questionnaire that assesses their understanding of others' intentions, as well as a questionnaire pertaining to their comprehension of social situations, ironic criticism, and idioms. Results: TD children outperformed the autistic children in idiom and irony understanding, as well as in ToM and social situation understanding. Understanding social situations and ToM contributed to idiom and irony understanding, with ToM ability uniquely contributing to irony (but not to idiom) understanding. Path analysis revealed that social cognitive abilities mediated the link between group affiliation and vocabulary, affecting figurative language comprehension. Conclusions: The present study demonstrates that social cognition skills influence the ability to understand ironic criticism and idioms, mediating the association between vocabulary and figurative language comprehension.

Research on Intelligent Mining Algorithm of English Translation Text Big Data Based on Deep Learning

This paper explores the translation of English texts by a deep learning algorithm, which aims to intelligently extract relevant records from a large corpus of English-translated texts, especially its shortcomings in processing difficult texts and dealing with long-term dependencies. This research integrates the modern long short-term memory (LSTM) network framework, with the aim of improving the accuracy and processing efficiency of the translation. Using a large number of experimental facts, these LSTM models are comprehensively learned and rigorously optimized. The traditional approach is to adapt the style to the language-specific properties of the translated text, and then use a bunch of evaluation metrics to evaluate the overall performance of the model across a large number of text types. The results of this study show that the method significantly speeds up the extraction and processing of information while maintaining the integrity at the translation level. In addition, the study reveals deep learning models’ ability to recognize complex contextual nuances and subtle formulations of language expression. The results of this research are multifaceted, bringing great improvements to device translation structures and facilitating the development of state-of-the-art text content analysis tools. These advances apply primarily to sectors of considerable statistical volume and complexity, including the criminal, scientific, and technological fields. Therefore, this study lays a foundation for future exploration of improving device understanding of language and automation of translation methods in various professional fields.

Exploring noise-induced techniques to strengthen deep learning model resilience

AbstractIn artificial intelligence, combating overfitting and enhancing model generalization is crucial. This research explores innovative noise-induced regularization techniques, focusing on natural language processing tasks. Inspired by gradient noise and Dropout, this study investigates the interplay between controlled noise, model complexity, and overfitting prevention. Utilizing long short-term memory and bidirectional long short term memory architectures, this study examines the impact of noise-induced regularization on robustness to noisy input data. Through extensive experimentation, this study shows that introducing controlled noise improves model generalization, especially in language understanding. This contributes to the theoretical understanding of noise-induced regularization, advancing reliable and adaptable artificial intelligence systems for natural language processing.

Computational Architecture of Speech Comprehension in the Human Brain

Understanding the computational algorithm that gives rise to human language is a shared endeavor among neuroscience, linguistics, and machine learning. We propose a conceptual framework for making measurable progress toward this goal by studying the subcomponents of the processing system: its underlying representations, operations, and information flow. We review evidence from neurophysiology, neuropsychology, linguistic theory, and computational modeling and suggest future directions to push the field forward in developing a precise characterization of spoken language understanding. Overall, we claim that representations of speech properties, and the operations that generate and manipulate those representations, exist within a highly parallel, highly redundant spatiotemporal regime.

Discourses of solidarity and resistance in alternative linguistics spaces: Galician improvised poetry as linguistic collective action

AbstractThis article focuses on alternative ways of understanding language in the context of minority language advocacy through an examination of the Galician tradition of singing‐in‐verse, known as regueifa. It proposes the notion of “linguistic collective action” to refer to the battery of resistance and solidarity strategies that lead to social transformation, implemented in grassroots movements linked to language struggles that not only go beyond the binary tropes of “pride” and “profit”, but also transcend traditional ideas of how language revitalization should be carried out. To tap into the dynamics of this social movement, we draw on a multi‐sited ethnography of the interconnected spaces in which urban‐based Galician speakers engage in collective action through the practice of regueifa. We examine how progressive values (e.g.: LGBTQ+ advocacy, feminism, and anti‐neoliberalism) are intertwined with the use of minoritized language reclamation, acting as a trigger for social transformation.

Testing and Evaluation of Health Care Applications of Large Language Models

Large language models (LLMs) can assist in various health care activities, but current evaluation approaches may not adequately identify the most useful application areas. To summarize existing evaluations of LLMs in health care in terms of 5 components: (1) evaluation data type, (2) health care task, (3) natural language processing (NLP) and natural language understanding (NLU) tasks, (4) dimension of evaluation, and (5) medical specialty. A systematic search of PubMed and Web of Science was performed for studies published between January 1, 2022, and February 19, 2024. Studies evaluating 1 or more LLMs in health care. Three independent reviewers categorized studies via keyword searches based on the data used, the health care tasks, the NLP and NLU tasks, the dimensions of evaluation, and the medical specialty. Of 519 studies reviewed, published between January 1, 2022, and February 19, 2024, only 5% used real patient care data for LLM evaluation. The most common health care tasks were assessing medical knowledge such as answering medical licensing examination questions (44.5%) and making diagnoses (19.5%). Administrative tasks such as assigning billing codes (0.2%) and writing prescriptions (0.2%) were less studied. For NLP and NLU tasks, most studies focused on question answering (84.2%), while tasks such as summarization (8.9%) and conversational dialogue (3.3%) were infrequent. Almost all studies (95.4%) used accuracy as the primary dimension of evaluation; fairness, bias, and toxicity (15.8%), deployment considerations (4.6%), and calibration and uncertainty (1.2%) were infrequently measured. Finally, in terms of medical specialty area, most studies were in generic health care applications (25.6%), internal medicine (16.4%), surgery (11.4%), and ophthalmology (6.9%), with nuclear medicine (0.6%), physical medicine (0.4%), and medical genetics (0.2%) being the least represented. Existing evaluations of LLMs mostly focus on accuracy of question answering for medical examinations, without consideration of real patient care data. Dimensions such as fairness, bias, and toxicity and deployment considerations received limited attention. Future evaluations should adopt standardized applications and metrics, use clinical data, and broaden focus to include a wider range of tasks and specialties.

Lexical Analysis of Japanese Children’s Songs from the Perspective of Japanese Language Education

Japanese children's songs can be a valuable resource for adult learners of the language, Firstly, the vocabulary used in children's songs is predominantly high-frequency and foundational, with a large portion falling below the N2 proficiency level. By mastering the words and phrases found in these songs, learners can expand their conversational abilities beyond the confines of traditional textbooks. Secondly, the songs offer a balanced representation of positive, negative, and neutral emotions, reflecting the emotional range and nuances of the Japanese language. This exposure to diverse emotional expressions can enrich learners' understanding of language and cultural nuances. Moreover, the study's semantic analysis reveals frequent word associations and co-occurrences, which can help learners grasp the interconnectedness of Japanese vocabulary and the cultural contexts in which words are used. Importantly, children's songs are deeply rooted in Japanese culture, indirectly exposing learners to various cultural aspects, such as the appreciation of nature, the concept of transience, and the importance of harmony and interconnectedness. This cultural immersion can foster a deeper understanding and appreciation of Japanese values and traditions. Furthermore, the familiar melodies and sing-along nature of these songs can facilitate social bonding and interpersonal connections among learners and native speakers. Engaging in shared cultural experiences, such as singing nursery rhymes together, can create a sense of community and strengthen the bond between learners and the target language culture.

Assessing the feasibility of ChatGPT-4o and Claude 3-Opus in thyroid nodule classification based on ultrasound images.

Large language models (LLMs) are pivotal in artificial intelligence, demonstrating advanced capabilities in natural language understanding and multimodal interactions, with significant potential in medical applications. This study explores the feasibility and efficacy of LLMs, specifically ChatGPT-4o and Claude 3-Opus, in classifying thyroid nodules using ultrasound images. This study included 112 patients with a total of 116 thyroid nodules, comprising 75 benign and 41 malignant cases. Ultrasound images of these nodules were analyzed using ChatGPT-4o and Claude 3-Opus to diagnose the benign or malignant nature of the nodules. An independent evaluation by a junior radiologist was also conducted. Diagnostic performance was assessed using Cohen's Kappa and receiver operating characteristic (ROC) curve analysis, referencing pathological diagnoses. ChatGPT-4o demonstrated poor agreement with pathological results (Kappa = 0.116), while Claude 3-Opus showed even lower agreement (Kappa = 0.034). The junior radiologist exhibited moderate agreement (Kappa = 0.450). ChatGPT-4o achieved an area under the ROC curve (AUC) of 57.0% (95% CI: 48.6-65.5%), slightly outperforming Claude 3-Opus (AUC of 52.0%, 95% CI: 43.2-60.9%). In contrast, the junior radiologist achieved a significantly higher AUC of 72.4% (95% CI: 63.7-81.1%). The unnecessary biopsy rates were 41.4% for ChatGPT-4o, 43.1% for Claude 3-Opus, and 12.1% for the junior radiologist. While LLMs such as ChatGPT-4o and Claude 3-Opus show promise for future applications in medical imaging, their current use in clinical diagnostics should be approached cautiously due to their limited accuracy.

Enhancing health assessments with large language models: A methodological approach.

Health assessments have long been a significant research topic within the field of health psychology. By analyzing the results of subject scales, these assessments effectively evaluate physical and mental health status. Traditional methods, based on statistical analysis, are limited in accuracy due to their reliance on linear scoring methods. Meanwhile, machine learning approaches, despite their potential, have not been widely adopted due to their poor interpretability and dependence on large amounts of training data. Recently, large language models (LLMs) have gained widespread attention for their powerful natural language understanding capabilities, offering a viable solution to these issues. This study investigates the application of LLMs in enhancing physical and mental health assessments, introducing ScaleLLM. ScaleLLM employs language and knowledge alignment to turn LLMs into expert evaluators for health psychology scales. Experimental results indicate that ScaleLLM can improve the accuracy and interpretability of health assessments.

Multimodal exploration of the thank God expressive construction and its implications for translation

Abstract Multimodal research in communication and translation studies is increasingly recognized, yet it remains incompletely explored. Leveraging computational linguistics with both Praat for acoustic analysis and the OpenPose and Rapid Annotator tools for visual analysis, this study delves into the intricate dynamics of the expressive construction thank God, providing a comprehensive examination of both visual and acoustic dimensions. Our objective is to uncover nuanced patterns of multimodal communication embedded within this expression and their implications for Translation and Interpreting. Through an analysis of linguistic features and co-speech gestures present in thank God, we aim to deepen our comprehension of how meaning crisscrosses modalities. Our findings underscore the necessity of a multimodal approach in language studies, emphasizing the requisite to preserve emotional and contextual nuances. The analysis unveils the phonological relevance of the duration of the construction’s second vowel, a key factor for translation. Additionally, data reveals a correlation between the emotion of relief and gestures executed with both hands closer to the chest. Overall, these findings contribute to advancing both multimodal communication research and translation studies, shedding light on the role of multimodal analysis in understanding language and translation dynamics, particularly in the context of constructions like thank God.

Joint intent detection and slot filling with syntactic and semantic features using multichannel CNN-BiLSTM

Understanding spoken language is crucial for conversational agents, with intent detection and slot filling being the primary tasks in natural language understanding (NLU). Enhancing the NLU tasks can lead to an accurate and efficient virtual assistant thereby reducing the need for human intervention and expanding their applicability in other domains. Traditionally, these tasks have been addressed individually, but recent studies have highlighted their interconnection, suggesting better results when solved together. Recent advances in natural language processing have shown that pretrained word embeddings can enhance text representation and improve the generalization capabilities of models. However, the challenge of poor generalization in joint learning models for intent detection and slot filling remains due to limited annotated datasets. Additionally, traditional models face difficulties in capturing both the semantic and syntactic nuances of language, which are vital for accurate intent detection and slot filling. This study proposes a hybridized text representation method using a multichannel convolutional neural network with three embedding channels: non-contextual embeddings for semantic information, part-of-speech (POS) tag embeddings for syntactic features, and contextual embeddings for deeper contextual understanding. Specifically, we utilized word2vec for non-contextual embeddings, one-hot vectors for POS tags, and bidirectional encoder representations from transformers (BERT) for contextual embeddings. These embeddings are processed through a convolutional layer and a shared bidirectional long short-term memory (BiLSTM) network, followed by two softmax functions for intent detection and slot filling. Experiments on the air travel information system (ATIS) and SNIPS datasets demonstrated that our model significantly outperformed the baseline models, achieving an intent accuracy of 97.90% and slot filling F1-score of 98.86% on the ATIS dataset, and an intent accuracy of 98.88% and slot filling F1-score of 97.07% on the SNIPS dataset. These results highlight the effectiveness of our proposed approach in advancing dialogue systems, and paving the way for more accurate and efficient natural language understanding in real-world applications.

Enhancing Large Language Model Comprehension of Material Phase Diagrams through Prompt Engineering and Benchmark Datasets

Large Language Models (LLMs) excel in fields such as natural language understanding, generation, complex reasoning, and biomedicine. With advancements in materials science, traditional manual annotation methods for phase diagrams have become inadequate due to their time-consuming nature and limitations in updating thermodynamic databases. To overcome these challenges, we propose a framework based on instruction tuning, utilizing LLMs for automated end-to-end annotation of phase diagrams. High-quality phase diagram images and expert descriptions are collected from handbooks and then preprocessed to correct errors, remove redundancies, and enhance information. These preprocessed data form a golden dataset, from which a subset are used to train LLMs through hierarchical sampling. The fine-tuned LLM is then tested for automated phase diagram annotation. Results show that the fine-tuned model achieves a cosine similarity of 0.8737, improving phase diagram comprehension accuracy by 7% compared to untuned LLMs. To the best of our knowledge, this is the first paper to propose using LLMs for the automated annotation of phase diagrams, replacing traditional manual annotation methods and significantly enhancing efficiency and accuracy.

Single and Multi-Hop Question-Answering Datasets for Reticular Chemistry with GPT-4-Turbo.

The rapid advancement in artificial intelligence and natural language processing has led to the development of large-scale datasets aimed at benchmarking the performance of machine learning models. Herein, we introduce "RetChemQA", a comprehensive benchmark dataset designed to evaluate the capabilities of such models in the domain of reticular chemistry. This dataset includes both single-hop and multi-hop question-answer pairs, encompassing approximately 45,000 question and answers (Q&As) for each type. The questions have been extracted from an extensive corpus of literature containing about 2,530 research papers from publishers including NAS, ACS, RSC, Elsevier, and Nature Publishing Group, among others. The dataset has been generated using OpenAI's GPT-4 Turbo, a cutting-edge model known for its exceptional language understanding and generation capabilities. In addition to the Q&A dataset, we also release a dataset of synthesis conditions extracted from the corpus of literature used in this study. The aim of RetChemQA is to provide a robust platform for the development and evaluation of advanced machine learning algorithms, particularly for the reticular chemistry community. The dataset is structured to reflect the complexities and nuances of real-world scientific discourse, thereby enabling nuanced performance assessments across a variety of tasks.

Sociolinguistic Style, Awareness, and Agency among Southern California Latinx Spanish–English Bilinguals

This study examined different degrees of awareness regarding the stigmatization of Southern California (SoCal) Spanish across four groups of Spanish–English bilinguals from Southern California (n = 87). The participants were presented with Spanish sentences and asked to decide which profile of speaker would likely express that sentence, given six options, such as: “someone living in Los Angeles/SoCal who grew up in Mexico” or “a Spanish-English bilingual who grew up in Los Angeles/SoCal”. Experimental stimuli included seven different linguistic categories of stigmatization, including English contact forms. The participants tended to attribute the stigmatized forms to bilinguals who grew up in Southern California. Central Colloquial and Taboo categories were more salient and perceived as forms used by people in Mexico. In contrast, English borrowings and redundancies seemed to be recognized by the participants, particularly for simultaneous bilinguals who grew up in Southern California, as salient forms of an identified Southern California Spanish variety. The results are interpreted within Exemplar Theory, with certain stigmatized forms indexing “Mexican Spanish” exemplars, and English borrowings identified as exemplars of SoCal Spanish. We advocate for usage-based approaches to understanding language perceptions and critical approaches to interrogating academic discourses.

Use of SNOMED CT in Large Language Models: Scoping Review.

Large language models (LLMs) have substantially advanced natural language processing (NLP) capabilities but often struggle with knowledge-driven tasks in specialized domains such as biomedicine. Integrating biomedical knowledge sources such as SNOMED CT into LLMs may enhance their performance on biomedical tasks. However, the methodologies and effectiveness of incorporating SNOMED CT into LLMs have not been systematically reviewed. This scoping review aims to examine how SNOMED CT is integrated into LLMs, focusing on (1) the types and components of LLMs being integrated with SNOMED CT, (2) which contents of SNOMED CT are being integrated, and (3) whether this integration improves LLM performance on NLP tasks. Following the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) guidelines, we searched ACM Digital Library, ACL Anthology, IEEE Xplore, PubMed, and Embase for relevant studies published from 2018 to 2023. Studies were included if they incorporated SNOMED CT into LLM pipelines for natural language understanding or generation tasks. Data on LLM types, SNOMED CT integration methods, end tasks, and performance metrics were extracted and synthesized. The review included 37 studies. Bidirectional Encoder Representations from Transformers and its biomedical variants were the most commonly used LLMs. Three main approaches for integrating SNOMED CT were identified: (1) incorporating SNOMED CT into LLM inputs (28/37, 76%), primarily using concept descriptions to expand training corpora; (2) integrating SNOMED CT into additional fusion modules (5/37, 14%); and (3) using SNOMED CT as an external knowledge retriever during inference (5/37, 14%). The most frequent end task was medical concept normalization (15/37, 41%), followed by entity extraction or typing and classification. While most studies (17/19, 89%) reported performance improvements after SNOMED CT integration, only a small fraction (19/37, 51%) provided direct comparisons. The reported gains varied widely across different metrics and tasks, ranging from 0.87% to 131.66%. However, some studies showed either no improvement or a decline in certain performance metrics. This review demonstrates diverse approaches for integrating SNOMED CT into LLMs, with a focus on using concept descriptions to enhance biomedical language understanding and generation. While the results suggest potential benefits of SNOMED CT integration, the lack of standardized evaluation methods and comprehensive performance reporting hinders definitive conclusions about its effectiveness. Future research should prioritize consistent reporting of performance comparisons and explore more sophisticated methods for incorporating SNOMED CT's relational structure into LLMs. In addition, the biomedical NLP community should develop standardized evaluation frameworks to better assess the impact of ontology integration on LLM performance.

Women’s challenges and gender inequality implications in the UK Home Office’s streamlined asylum process

This conceptual article scrutinises the challenges faced by women seeking asylum in the UK, focusing on the impact of the Home Office’s streamlined asylum process. Despite mention of gender-based issues in official guidelines, the system exhibits hostility and inadequacy in addressing the unique obstacles women encounter. Through a posthuman, practice-based theoretical framework, the article examines recent gender-based concerns raised about the Home Office’s use of questionnaires, machine translation, and informal linguistic assistance in lieu of interpreted asylum interviews for selected nationalities. It also highlights the implications of technologisation and bureaucratisation on linguistic access, arguing that the revised asylum-seeking process not only reinforces women’s marginalised position but also diminishes crucial resources like interpreting, in turn contributing to gendered power asymmetries between asylum seekers and the justice system. The article contends that this approach transfers the burden of language understanding to women, contributing to a climate of sociomaterial exclusion that curtails the rights of female asylum seekers to be adequately heard.

GAIA therapeutic farm: Prospective study on multidisciplinary care efficacy.

The therapeutic farm "GAIA" in Sidi Thabet is a non-profit organization created in 2009, situated in the North-Western suburbs of the capital Tunis/Tunisia. Its main aims are the day-care for and the socio-professional integration of children and young adults between age 6 and 30 with diverse types of disability through multiple activities, such as pet assisted therapy and horticulture. to assess the potential benefits of therapeutic on-farm management. We conducted a prospective study including 50 children and young adults aged 7 to 30 years with intellectual disabilities and/or autism spectrum disorders. Study participants were evaluated before inclusion and after 10 months of care using a self-established questionnaire collecting socio-demographic and clinical variables, an evaluation of competences, and the Child Behavior Checklist (CBCL). Mean sub scores (externalization/internalization) of CBCL decreased significantly (p<0.002). Competences were significantly improved in the following domains: understanding of verbal language, gross and fine motor skills, body hygiene, spatial and temporal orientation, preschool and school learning achievements and social integration. These positive initial results should encourage the further development of this initiative with support of the community.

6370 New-Onset Autoimmune Diabetes Following COVID-19 Infection and its Implications for Identical Twin Siblings

Abstract Disclosure: G. Wintermyer: None. S. Gillis-Funderburk: None. COVID-19 patients face many post-recovery complications,including autoimmune disorders. Late autoimmune diabetes ofadults (LADA) is one such disorder. Studies suggest that SARS-CoV may trigger proinflammatory cytokines that damage beta cellsand impair insulin secretion. Here, we explore a case of new-onsetLADA in a patient after recovery from COVID-19. Further, asidentical twins of LADA patients have a higher risk of developingit, we discuss how we approached the patient’s identical twin.A 29-year-old female with history of obesity (BMI 32.3) presentedto our endocrinology clinic. She complained of difficulty losingweight and fatigue and reported having COVID-19 three monthsearlier. She also mentioned having polyuria. She underwent labtests, including CBC, CMP, TSH, A1C, and a lipid panel. The testresults were within normal limits, except for her A1C of 11.3%.She was started on metformin and weekly Semaglutide injections.Given her recent COVID-19 history, she was assessed for LADA.Anti-GAD-65 and ZNT8 antibodies were 120 U/mL and 150U/mL, respectively. Serum C-peptide and IA-antibodies werewithin normal ranges. The patient informed us of her twin sister’sconcurrent COVID-19 infection. Due to the risk of LADA in hertwin sister, we recommended an evaluation by our clinic. Duringthe sister’s first visit, her BMI was 37.5, and she asked for weightloss medication. Comprehensive testing, including LADAantibodies, was conducted, with results within normal ranges andA1C level of 4.9%. The sister was also started on Semaglutide forweight management.The patient had quarterly follow-up appointments at the clinic.After a year, her A1C and continuous glucose monitoring machine(CGM) readings showed improvement. Her BMI decreased to 24.She never required insulin therapy, as her diabetes was well-managed and her average monthly glucose level on GCM readings were 125, 123 and 119 mg/dL in three different periods. Antibodieswere retested. This time IA-2 antibody was also elevated.Similarly, her twin sister continued to follow-up at our clinic. Thesister achieved significant weight loss without complications. HerLADA antibodies remained negative, and her A1C remainednormal. Clinicians should be vigilant for post-COVID LADA. In this case,Semaglutide aided weight loss, which benefits diabetesmanagement. Ongoing studies explore the potential of GLP-1medications in early-stage LADA. Historically, LADA patientswere started on insulin. However, CGM has enabled close glucosemonitoring, offering management of LADA with non-insulinmedications. It is also important to remember that autoimmunedisease risk is elevated in identical twins. In this case, although thetwin sister did not develop antibodies, the risk remains. Regularevaluation of high-risk patients is crucial for early detection ofantibodies, facilitating timely treatment to prevent adverseoutcomes. Presentation: 6/3/2024