Exploring The Effectiveness of In-Context Methods in Human-Aligned Large Language Models Across Languages

Large Language Models (LLMs) exhibit impressive zero/few-shot inference and generation quality for high-resource languages (HRLs).A few of them have been trained on low-resource languages (LRLs) and give decent performance.Owing to the prohibitive costs of training LLMs, they are usually used as a network service, with the client charged by the count of input and output tokens.The number of tokens strongly depends on the script and language, as well as the LLM's subword vocabulary.We show that LRLs are at a pricing disadvantage, because the wellknown LLMs produce more tokens for LRLs than HRLs.This is because most currently popular LLMs are optimized for HRL vocabularies.Our objective is to level the playing field: reduce the cost of processing LRLs in contemporary LLMs while ensuring that predictive and generative qualities are not compromised.As means to reduce the number of tokens processed by the LLM, we consider code-mixing, translation, and transliteration of LRLs to HRLs.We perform an extensive study using the IndicXTREME classification and six generative tasks dataset, covering 15 Indic and 3 other languages, while using GPT-4 (one of the costliest LLM services released so far 1 ) as a commercial LLM.We observe and analyze interesting patterns involving token count, cost, and quality across a multitude of languages and tasks.We show that choosing the best policy to interact with the LLM can reduce cost by 90% while giving better or comparable performance, compared to communicating with the LLM in the original LRL.

Mathematical Problem Solving in Arabic: Assessing Large Language Models

Large language models (LLMs) have demonstrated remarkable potential in handling multilingual machine translation (MMT).In this paper, we systematically investigate the advantages and challenges of LLMs for MMT by answering two questions: 1) How well do LLMs perform in translating massive languages?2) Which factors affect LLMs' performance in translation?We thoroughly evaluate eight popular LLMs, including ChatGPT and GPT-4.Our empirical results show that translation capabilities of LLMs are continually involving.GPT-4 has beat the strong supervised baseline NLLB in 40.91% of translation directions but still faces a large gap towards the commercial translation system like Google Translate, especially on low-resource languages.Through further analysis, we discover that LLMs exhibit new working patterns when used for MMT.First, LLM can acquire translation ability in a resource-efficient way and generate moderate translation even on zero-resource languages.Second, instruction semantics can surprisingly be ignored when given in-context exemplars.Third, cross-lingual exemplars can provide better task guidance for low-resource translation than exemplars in the same language pairs 1 .

The paper studies the problem of software code migration during the transition from the old version of the SAP ERP 6.0 system to modern ERP platforms, in particular SAP S/4HANA. The necessity of automation of the migration process using artificial intelligence methods and tools is substantiated. A comprehensive analysis of existing academic and industrial case studies confirmed the feasibility of using large language models for code migration. A number of advantages of the approach based on large language models are revealed, in particular, the ability to handle various migration scenarios. The key risks and limitations common in the enterprise environment are highlighted, including data security, privacy, high cost of API calls and limited knowledge of models regarding the subject area. A method is developed for selecting a large open source language model that can be safely used in the enterprise landscape, thereby mitigating the risks associated with cloud solutions. The method is based on the developed system of criteria, including: model size, context window size, license type, support for fine-tuning training, focus on code quality (measured using a reference dataset), support for model quantization and community support. The hybrid AHP-TOPSIS method is proposed for systematic ranking of candidate models. The AHP method was used to check the weights of the criteria, while TOPSIS was used to rank the models based on their proximity to the ideal solution. Based on the developed method, three large language models were selected: Qwen 2.5 Coder 14B, DeepSeek-Coder-V2 16B and Llama 3.1 8B. A comprehensive testing plan for the selected models was developed for further experimental studies. The scientific novelty of the work lies in the development of a method for selecting large language models for custom code migration tasks, which differs from the existing system of model selection criteria and the use of a hybrid AHP-TOPSIS approach for ranking candidate models. The practical value of the work is that the developed method allows you to select a model taking into account the constraints of the corporate environment and information security requirements, which will increase the level of automation in the migration of AVAP when switching to new versions of ERP systems from SAP SE. The developed method can be applied to the selection of large language models when upgrading any large computer systems developed in both open source and proprietary programming languages.

BackgroundDisease name recognition is a fundamental task in clinical natural language processing, enabling the extraction of critical patient information from electronic health records. While recent advances in large language models (LLMs) have shown promise, most evaluations have focused on English, and little is known about their robustness in low-resource languages such as Japanese. In particular, whether these models can perform reliably on previously unseen in-hospital data, which differs from training data in writing styles and clinical contexts, has not been thoroughly investigated.ObjectiveThis study evaluated the robustness of fine-tuned LLMs for disease name recognition in Japanese clinical notes, with a particular focus on their performance on in-hospital data that was not included during training.MethodsWe used two corpora for this study: (1) a publicly available set of Japanese case reports denoted as CR, and (2) a newly constructed corpus of progress notes, denoted as PN, written by ten physicians to capture stylistic variations of in-hospital clinical notes. To reflect real-world deployment scenarios, we first fine-tuned models on CR. Specifically, we compared a LLM and a baseline-masked language model (MLM). These models were then evaluated under two conditions: (1) on CR, representing the in-domain (ID) setting with the same document type, similar to training, and (2) on PN, representing the out-of-domain (OOD) setting with a different document type. Robustness was assessed by calculating the performance gap (ie, the performance drop from in-domain to out-of-domain settings).ResultsThe LLM demonstrated greater robustness, with a smaller performance gap in F1-scores (ID–OOD = −8.6) compared to the MLM baseline performance (ID–OOD = −13.9). This indicated more stable performance across ID and OOD settings, highlighting the effectiveness of fine-tuned LLMs for reliable use in diverse clinical settings.ConclusionsFine-tuned LLMs demonstrate superior robustness for disease name recognition in Japanese clinical notes, with a smaller performance gap. These findings highlight the potential of LLMs as reliable tools for clinical natural language processing in low-resource language settings and support their deployment in real-world health care applications, where diversity in documentation is inevitable.

Over the past few years, Large Language Models of Code (Code LLMs) have started to have a significant impact on programming practice. Code LLMs are also emerging as building blocks for research in programming languages and software engineering. However, the quality of code produced by a Code LLM varies significantly by programming language. Code LLMs produce impressive results on high-resource programming languages that are well represented in their training data (e.g., Java, Python, or JavaScript), but struggle with low-resource languages that have limited training data available (e.g., OCaml, Racket, and several others). This paper presents an effective approach for boosting the performance of Code LLMs on low-resource languages using semi-synthetic data. Our approach, called M ulti PL-T, generates high-quality datasets for low-resource languages, which can then be used to fine-tune any pretrained Code LLM. M ulti PL-T translates training data from high-resource languages into training data for low-resource languages in the following way. 1) We use a Code LLM to synthesize unit tests for commented code from a high-resource source language, filtering out faulty tests and code with low test coverage. 2) We use a Code LLM to translate the code from the high-resource source language to a target low-resource language. This gives us a corpus of candidate training data in the target language, but many of these translations are wrong. 3) We use a lightweight compiler to compile the test cases generated in (1) from the source language to the target language, which allows us to filter our obviously wrong translations. The result is a training corpus in the target low-resource language where all items have been validated with test cases. We apply this approach to generate tens of thousands of new, validated training items for five low-resource languages: Julia, Lua, OCaml, R, and Racket, using Python as the source high-resource language. Furthermore, we use an open Code LLM (StarCoderBase) with open training data (The Stack), which allows us to decontaminate benchmarks, train models without violating licenses, and run experiments that could not otherwise be done. Using datasets generated with M ulti PL-T, we present fine-tuned versions of StarCoderBase and Code Llama for Julia, Lua, OCaml, R, and Racket that outperform other fine-tunes of these base models on the natural language to code task. We also present Racket fine-tunes for two very recent models, DeepSeek Coder and StarCoder2, to show that M ulti PL-T continues to outperform other fine-tuning approaches for low-resource languages. The M ulti PL-T approach is easy to apply to new languages, and is significantly more efficient and effective than alternatives such as training longer.

Information Retrieval systems are becoming increasingly multimodal, and one interesting and natural modality for human information interaction is audio. Audio data is rapidly expanding in both volume and popularity. For example, more than 4 million podcasts are available worldwide - each containing tens to thousands of episodes - with over 500 million expected listeners as of 2024. Searching such large and evolving audio corpora involves multiple challenges, ranging from content and style diversity, expensive audio processing, and various indexing considerations such as retrieval segmentation and text vs audio retrieval modalities. In this thesis, we aim to address the challenges linked to audio retrieval and devise novel approaches to improve state-of-the-art (SOTA) performance. We divide the work into four main milestones that target answering the following research questions: RQ1: How does the quality of an ASR model affect retrieval effectiveness? One of the approaches to search spoken audio is to reduce the problem to text retrieval using ASR models to generate text transcripts. Spina et al. [2] showed that the quality of ASR has a clear impact on the summaries generated by ASR. Motivated by this, we are interested in examining how automatic transcription algorithms impact the effectiveness of audio retrieval and re-ranking in dense and sparse retrieval configurations. In many real-world scenarios, resource limits require cheaper ASR models to be used. Therefore, we are interested in exploring if dense semantic search methods can reduce the impact of ASR model choice. RQ2: How can we build more complete, comprehensive, and representative audio retrieval resources to promote further research on audio retrieval? After reviewing the literature and performing a preliminary experiment on the Spotify podcasts dataset [1], we noted two main gaps in the available resources. Firstly, the large audio dataset released in TREC suffers from the shallow pooling issue, where new systems retrieve many unseen judgments, making it difficult to assess the quality of modern IR techniques. Secondly, there is a scarcity of audio datasets available for low-resource languages. Our plan to bridge these gaps comprises two parts: (1) Assessing the missed judgments Motivated by the recent literature on using LLMs as a judge [3], we plan to resolve the unjudged passages issue by assessing the relevance with multiple large language models (LLMs). (2) Curate a multi-lingual audio dataset with a focus on low-resource languages. Since there is a scarcity of audio datasets in low-resource languages, we plan to collect a large number of audio files in multiple low-resource languages, unify their format, and make them publicly available as a first step toward building test collections that enable research and advancements in such languages. RQ3: How can we combine audio and text to produce an efficient and effective multi-modal ranking model? We will focus on solutions that depend on processing both text and audio signals during indexing and retrieval. Our first study will be focused on improving the effectiveness of low-cost ASR systems. Another promising direction is to use the phonetic features as a representation of a word or token. Then, we can devise a suitable indexing paradigm to support searching over these representations. RQ4: What are the challenges affecting audio indexing and search performance, and what is the best way to mitigate against similar issues in the future? One of the main challenges in multi-modal search is identifying the suitable length for the retrieval unit. The retrieval unit satisfying an information need might vary in length from a couple of minutes to 10 minutes to one hour or a whole episode or series. Other challenges will stem from failure cases within state-of-the-art text/audio retrieval systems. So, in parallel with answering the previous questions, we will also study such failure cases, investigate the reasons behind them, and propose solutions that can mitigate their negative impacts.

Background Autoimmune hepatitis (AIH) is a chronic immune-mediated liver disease that requires long-term management, in which effective patient education plays a critical role. With the rapid development of large language models (LLMs), AI-generated health information is increasingly accessed by patients; however, the readability, quality, and educational suitability of LLM-generated AIH-related content remain insufficiently evaluated. Methods Five widely used LLMs—ChatGPT, Doubao, DeepSeek, Wenxin Yiyan, and Tongyi Qianwen—were assessed based on their responses to 20 frequently asked AIH patient education questions covering five thematic categories. Text readability was evaluated using multiple indices, including the Automated Readability Index, Flesch Reading Ease Score, Gunning Fog Index, Flesch–Kincaid Grade Level, Coleman–Liau Index, SMOG, and Linsear Write formula. Information quality and educational suitability were assessed using the Global Quality Score (GQS) and the Chinese version of the Patient Education Materials Assessment Tool (C-PEMAT). Clinical Intent Alignment (CIA) was used to evaluate the coverage of guideline-defined medical key points based on the 2025 EASL Clinical Practice Guidelines. Inter-rater reliability was analyzed using Cohen’s kappa, and comparative and correlation analyses were performed. Results Significant differences were observed among the LLMs in readability, information quality, and educational suitability (all p < 0.05). ChatGPT achieved the highest GQS and C-PEMAT scores, followed by Doubao and DeepSeek, whereas Wenxin Yiyan and Tongyi Qianwen showed lower performance and greater variability. CIA analysis indicated comparable coverage of guideline-defined clinical intent across models. Readability varied significantly across content themes, with texts related to disease mechanisms and diagnostic processes exhibiting higher linguistic complexity. Correlation analysis demonstrated moderate associations between GQS and grade-level readability indices, whereas C-PEMAT and CIA showed weak correlations with traditional readability metrics. Conclusion Substantial variability exists among LLMs in generating AIH patient education materials. Model selection critically influences information quality and educational suitability, whereas content theme primarily affects linguistic complexity. Although most models produced moderate-to-good quality information, relatively high readability levels suggest that further simplification may be needed for general patient populations. A multidimensional evaluation framework integrating readability, quality, educational suitability, and clinical intent alignment is essential for the responsible use of LLMs in AIH patient education.

Multi-Language Speech-to-Text Translation (ST) plays a pivotal role in bridging linguistic barriers by converting spoken language into written text across different languages.This project aims to develop a robust ST model tailored for low-resource Indian languages, specifically targeting the Indo-Aryan and Dravidian language families.The dataset used consists of speeches from conferences and TED Talks, along with their corresponding transcriptions in English (source language) and translations in Hindi, Bengali, and Tamil (target languages).By tackling the lack of data and disparities in attention within low-resource languages, the paper strives to create an efficient ST system capable of real-world deployment.Additionally, existing resources in related languages are leveraged and word-level translation resources are explored to enhance translation accuracy.

Generative large language models (LLMs) are a subset of transformers-based neural network architecture models. LLMs have successfully leveraged a combination of an increased number of parameters, improvements in computational efficiency, and large pre-training datasets to perform a wide spectrum of natural language processing (NLP) tasks. Using a few examples (few-shot) or no examples (zero-shot) for prompt-tuning has enabled LLMs to achieve state-of-the-art performance in a broad range of NLP applications. This article by the American Medical Informatics Association (AMIA) NLP Working Group characterizes the opportunities, challenges, and best practices for our community to leverage and advance the integration of LLMs in downstream NLP applications effectively. This can be accomplished through a variety of approaches, including augmented prompting, instruction prompt tuning, and reinforcement learning from human feedback (RLHF). Our focus is on making LLMs accessible to the broader biomedical informatics community, including clinicians and researchers who may be unfamiliar with NLP. Additionally, NLP practitioners may gain insight from the described best practices. We focus on 3 broad categories of NLP tasks, namely natural language understanding, natural language inferencing, and natural language generation. We review the emerging trends in prompt tuning, instruction fine-tuning, and evaluation metrics used for LLMs while drawing attention to several issues that impact biomedical NLP applications, including falsehoods in generated text (confabulation/hallucinations), toxicity, and dataset contamination leading to overfitting. We also review potential approaches to address some of these current challenges in LLMs, such as chain of thought prompting, and the phenomena of emergent capabilities observed in LLMs that can be leveraged to address complex NLP challenge in biomedical applications.

Generative Artificial Intelligence (GenAI), particularly Large Language Models (LLMs), has significantly advanced Natural Language Processing (NLP) tasks, such as Named Entity Recognition (NER), which involves identifying entities like person, location, and organization names in text. LLMs are especially promising for low-resource languages due to their ability to learn from limited data. However, the performance of GenAI models for Nepali, a low-resource language, has not been thoroughly evaluated. This paper investigates the application of state-of-the-art LLMs for Nepali NER, conducting experiments with various prompting techniques to assess their effectiveness. Our results provide insights into the challenges and opportunities of using LLMs for NER in low-resource settings and offer valuable contributions to the advancement of NLP research in languages like Nepali.

Code comment generation typically refers to the process of generating concise natural language descriptions for a piece of code, which facilitates program comprehension activities. Inline code comments, as a part of code comments, are also crucial for program comprehension. Recently, the emergence of large language models (LLMs) has significantly boosted the performance of natural language processing tasks. This naturally inspires us to explore the performance of the LLMs in the task of inline code comment generation. To this end, we evaluate open-source LLMs on a large-scale dataset and compare the results with the current state-of-the-art methods. Specifically, we explore the model performance in the following scenarios based on the widely used evaluation metrics (i.e. BLEU, Meteor, and ROUGE-L): (1) generation with simple instruction; (2) few-shot-guided generation with random examples selected from the database; (3) few-shot-guided generation with similar examples selected from the database; and (4) adopt the re-ranking strategy for the output of LLMs. Our findings reveal that: (1) under the simple instruction scenario, LLMs could not fully show the potential in the task of inline comment generation compared to the state-of-the-art models; (2) random few-shot leads to a slight improvement; (3) similar few-shot and re-ranking strategy could significantly enhance the performance of LLMs; and (4) for inline comment and code snippet pairs with different intents, why category achieves the best performance and what category achieves relatively poorer performance. That remains consistent across all four scenarios. Our findings shed light on future research directions for using LLMs in inline comment generation tasks.

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 paper studies the performance of open-source Large Language Models (LLMs) in text classification tasks typical for political science research. By examining tasks like stance, topic, and relevance classification, we aim to guide scholars in making informed decisions about their use of LLMs for text analysis and to establish a baseline performance benchmark that demonstrates the models’ effectiveness. Specifically, we conduct an assessment of both zero-shot and fine-tuned LLMs across a range of text annotation tasks using news articles and tweets datasets. Our analysis shows that fine-tuning improves the performance of open-source LLMs, allowing them to match or even surpass zero-shot GPT-3.5 and GPT-4, though still lagging behind fine-tuned GPT-3.5. We further establish that fine-tuning is preferable to few-shot training with a relatively modest quantity of annotated text. Our findings show that fine-tuned open-source LLMs can be effectively deployed in a broad spectrum of text annotation applications. We provide a Python notebook facilitating the application of LLMs in text annotation for other researchers.

Game theory establishes a fundamental framework for analyzing strategic interactions among rational decision-makers. The rapid advancement of large language models (LLMs) has sparked extensive research exploring the intersection of these two fields. Specifically, game-theoretic methods are being applied to evaluate and enhance LLM capabilities, while LLMs themselves are reshaping classic game models. This paper presents a comprehensive survey of the intersection of these fields, exploring a bidirectional relationship from three perspectives: (1) Establishing standardized game-based benchmarks for evaluating LLM behavior; (2) Leveraging game-theoretic methods to improve LLM performance through algorithmic innovations; (3) Characterizing the societal impacts of LLMs through game modeling. Among these three aspects, we also highlight how the equilibrium analysis for traditional game models is impacted by LLMs' advanced language understanding, which in turn extends the study of game theory. Finally, we identify key challenges and future research directions, assessing their feasibility based on the current state of the field. By bridging theoretical rigor with emerging AI capabilities, this survey aims to foster interdisciplinary collaboration and drive progress in this evolving research area.