Abstract While extensive research has identified the drivers of algorithm aversion, the influence of users’ subjective perceptions remains largely underexplored. This study investigates the psychological mechanisms linking perceived algorithmic ineffectiveness to algorithm aversion behavior, proposing psychological reactance as a mediator and implicit personality as a moderator. Data were collected from 733 users of a Chinese language short-form video platform called Douyin, known as TikTok in other regions. The results show that (1) perceived algorithmic ineffectiveness positively predicts algorithm aversion behavior, (2) psychological reactance partially mediates this relationship, and (3) implicit personality moderates the link between perceived algorithmic ineffectiveness and psychological reactance, with the effect being stronger for incremental theorists than for entity theorists. These findings establish perceived algorithmic ineffectiveness as a key driver of algorithm aversion, contributing to the current literature on the psychological underpinnings of algorithm aversion behavior.
History of scholarship and learning. The humanities, Social Sciences
Multimodal research has predominantly focused on single-image reasoning, with limited exploration of multi-image scenarios. Recent models have sought to enhance multi-image understanding through large-scale pretraining on interleaved image-text datasets. However, most Vision-Language Models (VLMs) are trained primarily on English datasets, leading to inadequate representation of Indian languages. To address this gap, we introduce the Chitrakshara dataset series, covering 11 Indian languages sourced from Common Crawl. It comprises (1) Chitrakshara-IL, a large-scale interleaved pretraining dataset with 193M images, 30B text tokens, and 50M multilingual documents, and (2) Chitrakshara-Cap, which includes 44M image-text pairs with 733M tokens. This paper details the data collection pipeline, including curation, filtering, and processing methodologies. Additionally, we present a comprehensive quality and diversity analysis to assess the dataset's representativeness across Indic languages and its potential for developing more culturally inclusive VLMs.
Coded language is an important part of human communication. It refers to cases where users intentionally encode meaning so that the surface text differs from the intended meaning and must be decoded to be understood. Current language models handle coded language poorly. Progress has been limited by the lack of real-world datasets and clear taxonomies. This paper introduces CodedLang, a dataset of 7,744 Chinese Google Maps reviews, including 900 reviews with span-level annotations of coded language. We developed a seven-class taxonomy that captures common encoding strategies, including phonetic, orthographic, and cross-lingual substitutions. We benchmarked language models on coded language detection, classification, and review rating prediction. Results show that even strong models can fail to identify or understand coded language. Because many coded expressions rely on pronunciation-based strategies, we further conducted a phonetic analysis of coded and decoded forms. Together, our results highlight coded language as an important and underexplored challenge for real-world NLP systems.
Eneko Valero, Maria Ribalta i Albado, Oscar Sainz
et al.
Large Language Models (LLMs) remain heavily centered on English, with limited performance in low-resource languages. Existing adaptation approaches, such as continual pre-training, demand significant computational resources. In the case of instructed models, high-quality instruction data is also required, both of which are often inaccessible for low-resource language communities. Under these constraints, model merging offers a lightweight alternative, but its potential in low-resource contexts has not been systematically explored. In this work, we explore whether it is possible to transfer language knowledge to an instruction-tuned LLM by merging it with a language-specific base model, thereby eliminating the need of language-specific instructions and repeated fine-tuning processes whenever stronger instructed variants become available. Through experiments covering four Iberian languages (Basque, Catalan, Galician, and Spanish) and two model families, we show that merging enables effective instruction following behavior in new languages and even supports multilingual capability through the combination of multiple language-specific models. Our results indicate that model merging is a viable and efficient alternative to traditional adaptation methods for low-resource languages, achieving competitive performance while greatly reducing computational cost.
The character “雜” (zá), commonly found in Chinese Buddhist literature, typically conveys the meaning of “mixed” or “varied”. However, in the translations of the renowned Tang dynasty translator Xuanzang, its usage stands out both in frequency and distinctiveness, setting his work apart from that of other translators. Terms traditionally conveyed using “不淨” (bù jìng, “impure”) or “穢” (huì, “filth”) were deliberately transformed by Xuanzang into “雜染” (zá rǎn, “mixed defilement”) and “雜穢” (zá huì, “mixed filth”), with “雜” nearly becoming synonymous with impurity. Examining the original meaning of “雜”, we find that it primarily signifies “to gather” or “miscellaneous”, typically carrying a neutral connotation. However, when used as an adjective describing a state, “雜” transcends its neutral sense of “various” or “diverse” to encompass notions of impurity, disorder, and deviation from normative standards—often with negative implications. Building on this understanding, it becomes clear that the abstract opposition between purity and impurity in the doctrinal meanings of Buddhist scriptures was reinterpreted by Xuanzang as a concrete opposition between “清淨” (qīng jìng, “purity”) and “雜穢” (mixed filth). This reinterpretation allowed “雜” to describe anything defiling the mind or carrying negative overtones—even when the original Sanskrit text did not explicitly indicate such a notion—thereby constituting a strategic substitution in translation. Furthermore, Xuanzang and his contemporaries frequently employed “雜” as a functional component within disyllabic compounds that collectively expressed negative meanings. Some terms containing “雜” thus cannot be understood simply as “mixed” or “varied”; instead, “雜” functions as a negative marker, reinforcing unfavorable connotations. This paper provides a focused case study on the lexical strategies of ancient Buddhist translators, illustrating how particular concepts—including 雜—were leveraged to reshape doctrinal content. In doing so, it highlights the deliberate linguistic and interpretative choices made by translators like Xuanzang, offering insights into their motivations and the cultural–linguistic contexts that framed their work.
Abstract Objectives This study aimed to synthesize existing evidence to compare the immunogenicity and safety of domestic and imported live-attenuated varicella vaccine (VarV) in healthy Chinese populations. Methods We searched PubMed, Web of Science, Embase, China National Knowledge Internet (CNKI), Wan Fang Database, and Chinese Biomedical Literature Service System (SinoMed) using predefined search terms to identify relevant studies. Retrieve all language articles up to March 15, 2024. Articles reported varicella vaccination in healthy Chinese populations were included. We calculated the pooled rates of seroconversions and adverse events using the random effects model and assessed the quality of each study using the modified Jadad Scale and Newcastle Ottawa Scale (NOS). Publication bias was evaluated using Egger’s regression test. Results In our immunogenicity analysis, which included 16,655 Chinese individuals from 21 studies, the pooled seroconversion rate was 89% (95%CI: 86-91%) for domestic VarV and 93% (95%CI: 88-98%) for imported Varv, with no statistically significant difference. In our safety analysis, which included 29,696 Chinese individuals from 25 studies, the pooled rate of systemic reactions was higher for domestic Varv (11%, 95%CI: 10-13%) than for imported Varv (8%, 95%CI: 6-10%; P < 0.001), while the results for local reactions were the opposite (domestic Varv: 3%, 95%CI: 2-3%; imported Varv: 7%, 95%CI: 3-10%; P = 0.020). The results are based on pooled proportions rather than direct comparison. Egger’s test suggested that publication bias was not negligible. Conclusions Both domestic and imported varicella vaccines appear to be generally immunogenic and safe in healthy Chinese populations. However, due to limited and heterogeneous data on imported vaccines, further high-quality studies are needed to validate these comparative findings.
As an important form of contemporary presentation of Oriental aesthetics in the design field, the neo-Chinese style has formed a unique cultural expression through the creative transformation and innovative development of traditional aesthetic elements. This paper deeply explores the three major cultural genes of neo-Chinese style product design——the artistic conception creation of "blank space", the vitality expression of "vitality and spirit", and the sense of order of "symmetry and balance", and systematically analyzes the modern translation paths of these traditional aesthetic elements in design practice. Research shows that neo-Chinese design has realized the organic integration of traditional cul-ture and modern design through a methodological system including element extraction and reconstruction, material innovation and craft integration, scene adaptation and function optimization. This paper further summarizes the opera-ble framework of neo-Chinese design methodology, and puts forward prospects for its cultural depth enhancement and digital technology integration path in future development, providing theoretical reference and practical guidance for contemporary design practice.
Automatic speech recognition systems have undoubtedly advanced with the integration of multilingual and multitask models such as Whisper, which have shown a promising ability to understand and process speech across a wide range of languages. Despite their robustness, these models often fall short in handling the linguistic distinctions of minority languages. This study addresses this gap by integrating traditional and novel language models with fine-tuned Whisper models to raise their performance in less commonly studied languages. Through rigorous fine-tuning and evaluation across multiple datasets, we demonstrate substantial improvements in word error rate, particularly in low-resource scenarios. Our approach not only does take advantage of the extensive data Whisper was pre-trained on, but also complements its linguistic adaptability by incorporating language models. We obtained improvements up to 51% for in-distribution datasets and up to 34% for out-of-distribution sentences using statistical language models, while large language models provided moderate but consistently robust improvement across diverse linguistic contexts. The findings reveal that, while the integration reliably benefits all model sizes, the extent of improvement varies, highlighting the importance of optimized language model parameters. Finally, we emphasize the importance of selecting appropriate evaluation parameters when reporting the results using transformer-based ASR models. In summary, this research clears the way for more inclusive ASR technologies that perform better across languages by enriching their linguistic knowledge. For further implementation details of this study, the technical documentation and source code are available at http://www.github.com/hitz-zentroa/whisper-lm.
Natural language-based assessment (NLA) is an approach to second language assessment that uses instructions - expressed in the form of can-do descriptors - originally intended for human examiners, aiming to determine whether large language models (LLMs) can interpret and apply them in ways comparable to human assessment. In this work, we explore the use of such descriptors with an open-source LLM, Qwen 2.5 72B, to assess responses from the publicly available S&I Corpus in a zero-shot setting. Our results show that this approach - relying solely on textual information - achieves competitive performance: while it does not outperform state-of-the-art speech LLMs fine-tuned for the task, it surpasses a BERT-based model trained specifically for this purpose. NLA proves particularly effective in mismatched task settings, is generalisable to other data types and languages, and offers greater interpretability, as it is grounded in clearly explainable, widely applicable language descriptors.
The evolution of large language models (LLMs) toward artificial superhuman intelligence (ASI) hinges on data reproduction, a cyclical process in which models generate, curate and retrain on novel data to refine capabilities. Current methods, however, risk getting stuck in a data reproduction trap: optimizing outputs within fixed human-generated distributions in a closed loop leads to stagnation, as models merely recombine existing knowledge rather than explore new frontiers. In this paper, we propose language games as a pathway to expanded data reproduction, breaking this cycle through three mechanisms: (1) \textit{role fluidity}, which enhances data diversity and coverage by enabling multi-agent systems to dynamically shift roles across tasks; (2) \textit{reward variety}, embedding multiple feedback criteria that can drive complex intelligent behaviors; and (3) \textit{rule plasticity}, iteratively evolving interaction constraints to foster learnability, thereby injecting continual novelty. By scaling language games into global sociotechnical ecosystems, human-AI co-evolution generates unbounded data streams that drive open-ended exploration. This framework redefines data reproduction not as a closed loop but as an engine for superhuman intelligence.
Recent research shows that punctuation patterns in texts exhibit universal features across languages. Analysis of Western classical literature reveals that the distribution of spaces between punctuation marks aligns with a discrete Weibull distribution, typically used in survival analysis. By extending this analysis to Chinese literature represented here by three notable contemporary works, it is shown that Zipf's law applies to Chinese texts similarly to Western texts, where punctuation patterns also improve adherence to the law. Additionally, the distance distribution between punctuation marks in Chinese texts follows the Weibull model, though larger spacing is less frequent than in English translations. Sentence-ending punctuation, representing sentence length, diverges more from this pattern, reflecting greater flexibility in sentence length. This variability supports the formation of complex, multifractal sentence structures, particularly evident in Gao Xingjian's "Soul Mountain". These findings demonstrate that both Chinese and Western texts share universal punctuation and word distribution patterns, underscoring their broad applicability across languages.
We introduce a novel paradigm of emergent local memory. It is a continuous-learning completely-parallel content-addressable memory encoding global order. It demonstrates how local constraints on uncoordinated learning can produce topologically protected memories realizing emergent symbolic order. It is therefore a neuro-symbolic bridge. It further has the ability to produce human language without data, by exploiting its own self-organizing dynamics. It teaches us that words arise as a side-effect of emergent symbolic order, and that human language patterns at all structural levels reflect a universal mechanism of word formation (which is subregular). This work answers essential questions about the existence \& origin of all the human language data.
The evaluation of factual accuracy in large vision language models (LVLMs) has lagged behind their rapid development, making it challenging to fully reflect these models' knowledge capacity and reliability. In this paper, we introduce the first factuality-based visual question-answering benchmark in Chinese, named ChineseSimpleVQA, aimed at assessing the visual factuality of LVLMs across 8 major topics and 56 subtopics. The key features of this benchmark include a focus on the Chinese language, diverse knowledge types, a multi-hop question construction, high-quality data, static consistency, and easy-to-evaluate through short answers. Moreover, we contribute a rigorous data construction pipeline and decouple the visual factuality into two parts: seeing the world (i.e., object recognition) and discovering knowledge. This decoupling allows us to analyze the capability boundaries and execution mechanisms of LVLMs. Subsequently, we evaluate 34 advanced open-source and closed-source models, revealing critical performance gaps within this field. Our evaluation-friendly code and data have already been open-sourced.
Cross-lingual Natural Language Processing (NLP) has gained significant traction in recent years, offering practical solutions in low-resource settings by transferring linguistic knowledge from resource-rich to low-resource languages. This field leverages techniques like annotation projection and model transfer for language adaptation, supported by multilingual pre-trained language models. However, linguistic divergences hinder language transfer, especially among typologically distant languages. In this paper, we present an analysis of predicate-argument structures in parallel Chinese and English sentences. We explore the alignment and misalignment of predicate annotations, inspecting similarities and differences and proposing a categorization of structural divergences. The analysis and the categorization are supported by a qualitative and quantitative analysis of the results of an annotation projection experiment, in which, in turn, one of the two languages has been used as source language to project annotations into the corresponding parallel sentences. The results of this analysis show clearly that language transfer is asymmetric. An aspect that requires attention when it comes to selecting the source language in transfer learning applications and that needs to be investigated before any scientific claim about cross-lingual NLP is proposed.
Multilingual large language models (LLMs) are known to more frequently generate non-faithful output in resource-constrained languages (Guerreiro et al., 2023 - arXiv:2303.16104), potentially because these typologically diverse languages are underrepresented in their training data. To mitigate unfaithfulness in such settings, we propose using computationally light auxiliary models to rescore the outputs of larger architectures. As proof of the feasibility of such an approach, we show that monolingual 4-layer BERT models pretrained from scratch on less than 700 MB of data without fine-tuning are able to identify faithful summaries with a mean accuracy of 88.33% in three genetically unrelated languages that differ in their morphological complexity - Vietnamese, Polish and Georgian. The same hyperparameter combination moreover generalises well to three other tasks, suggesting applications for rescoring beyond improving faithfulness. In order to inform typologically aware model selection, we also investigate how morphological complexity interacts with regularisation, model depth and training objectives, ultimately demonstrating that morphologically complex languages are more likely to benefit from dropout, while across languages downstream performance is enhanced most by shallow architectures as well as training using the standard BERT objectives.
The rapid advancement of large language models (LLMs) has reshaped the landscape of machine translation, yet challenges persist in preserving poetic intent, cultural heritage, and handling specialized terminology in Chinese-English translation. This study constructs a diverse corpus encompassing Chinese scientific terminology, historical translation paradoxes, and literary metaphors. Utilizing a back-translation and Friedman test-based evaluation system (BT-Fried), we evaluate BLEU, CHRF, TER, and semantic similarity metrics across six major LLMs (e.g., GPT-4.5, DeepSeek V3) and three traditional translation tools. Key findings include: (1) Scientific abstracts often benefit from back-translation, while traditional tools outperform LLMs in linguistically distinct texts; (2) LLMs struggle with cultural and literary retention, exemplifying the "paradox of poetic intent"; (3) Some models exhibit "verbatim back-translation", reflecting emergent memory behavior; (4) A novel BLEU variant using Jieba segmentation and n-gram weighting is proposed. The study contributes to the empirical evaluation of Chinese NLP performance and advances understanding of cultural fidelity in AI-mediated translation.
Abstract Background: Posttraumatic headache (PTH) after traumatic brain injury (TBI) is a common clinical symptom, which refers to a headache that occurs after TBI. Acupuncture is often used for the treatment of such patients in China, and significant clinical effects have been achieved. However, to date, its efficacy has not been methodically evaluated. The purpose of this systematic review is to provide evidence to prove the effectiveness of acupuncture in the treatment of PTH in patients with TBI. Methods: This systematic review will be conducted in accordance with the preferred reporting items for systematic review and meta-analysis protocols. The following electronic databases will be searched from their inception to February 2022: PubMed, Web of Science, Embase, PsycINFO, the Cochrane Library, and Chinese databases such as Chinese Biomedical Literature (CBM), Chinese Medical Current Content (CMCC), Chinese Scientific Journal Database (VIP), WanFang Database, and China National Knowledge Infrastructure (CNKI). No language restrictions will be applied to the search strategy. Randomized controlled trials and cohort and case-control studies that met the inclusion and exclusion criteria will be included in this study. The meta-analysis will be performed using RevMan 5.3 software. Each session of this systematic review will be conducted independently by 2 members. Results: This review evaluates the efficacy of acupuncture in the treatment of PTH after TBI. Conclusion: This review provides substantial evidence for the clinical application of acupuncture in PTH treatment after TBI. Ethics and dissemination: Since the data in this study will be retrieved from published trials, therefore the Patient Consent Statement and Ethical Approval are not required. We will disseminate our results by publishing the research in a peer-reviewed journal. Trail registration number: The protocol was registered in INPLASY (INPLASY 202220073).
This paper explores how to use ChatGPT-4 as an assistant for personalized teaching in the poetry writing class of Class 1 of the third-year university majoring in Chinese language and literature, to realize the human-computer collaboration between ChatGPT and the teacher, and conduct personalized teaching to provide students with personalized teaching. professional guidance and timely feedback. Traditional poetry writing teaching is difficult to meet the personalized needs of large-scale classes, and the introduction of artificial intelligence (AI) assistants can effectively solve this challenge. Therefore, this paper discusses how to use the advice of ChatGPT -4 and the professional guidance of teachers to improve the efficiency and quality of students' poetry homework correction, achieve personalized teaching guidance for students, and thus promote the improvement of poetry writing skills.
As people increasingly prioritize their health, the speed and breadth of health information dissemination on the internet have also grown. At the same time, the presence of false health information (health rumors) intermingled with genuine content poses a significant potential threat to public health. However, current research on Chinese health rumors still lacks a large-scale, public, and open-source dataset of health rumor information, as well as effective and reliable rumor detection methods. This paper addresses this gap by constructing a dataset containing 1.12 million health-related rumors (HealthRCN) through web scraping of common health-related questions and a series of data processing steps. HealthRCN is the largest known dataset of Chinese health information rumors to date. Based on this dataset, we propose retrieval-augmented large language models for Chinese health rumor detection and explainability (HRDE). This model leverages retrieved relevant information to accurately determine whether the input health information is a rumor and provides explanatory responses, effectively aiding users in verifying the authenticity of health information. In evaluation experiments, we compared multiple models and found that HRDE outperformed them all, including GPT-4-1106-Preview, in rumor detection accuracy and answer quality. HRDE achieved an average accuracy of 91.04% and an F1 score of 91.58%.
Developmental psychologists have argued about when cognitive capacities such as language understanding or theory of mind emerge. These debates often hinge on the concept of "task demands" -- the auxiliary challenges associated with performing a particular evaluation -- that may mask the child's underlying ability. The same issues arise when measuring the capacities of language models (LMs): performance on a task is a function of the model's underlying knowledge, combined with the model's ability to interpret and perform the task given its available resources. Here, we show that for analogical reasoning, reflective reasoning, word prediction, and grammaticality judgments, evaluation methods with greater task demands yield lower performance than evaluations with reduced demands. This "demand gap" is most pronounced for models with fewer parameters and less training data. Our results illustrate that LM performance should not be interpreted as a direct indication of intelligence (or lack thereof), but as a reflection of capacities seen through the lens of researchers' design choices.