Small Language Models (SLMs) enable cost - effective, on - device and latency - sensitive AI applications, yet their deployment in Traditional Chinese (TC) remains hindered by token - level instability -- models unpredictably emit non - TC characters or code - switch into othe r languages. We address this practical reliability gap by creating PureTC - 1B, a three - stage stabilization pipeline for Llama - 3.2 - 1B - Instruct (an open - weight, instruction - tuned model released by Meta) [1] using parameter - efficient LoRA adapters [2] . Our met hod combines Continual Pre - Training (CPT) on TC - centric corpora, Supervised Fine - Tuning (SFT) with instruction data, and Direct Preference Optimization (DPO) [3] using TC - adherence preferences to improve monolingual robustness without full - model retraining. On a benchmark designed to simulate real - world usage, PureTC - 1B achieves a 51.3% relative reduction (micro - average) in non - TC output tokens versus the base model. On a Named Entity Translation (NET) task, PureTC - 1B further reduces incorrect - language tokens by 77.2% relative to Llama - 3B and 57.2% relative to Qwen - 1.5B, indicating that robust 2 of 17 TC adherence is attainable even at the 1B scale. The pipeline is reproducible, adapter - only, and hardware - friendly, offering practitioners a practical recipe to enhance language stability for TC and potentially other non - English languages.

Multimodal Large Language Models (MLLMs) process visual, acoustic, and textual inputs, addressing the limitations of single-modality LLMs. However, existing benchmarks often overlook tri-modal evaluation in Traditional Chinese and do not consider inference latency. To address this, we introduce Multi-TW, the first Traditional Chinese benchmark for evaluating the performance and latency of any-to-any multimodal models. Multi-TW includes 900 multiple-choice questions (image and text, audio and text pairs) sourced from official proficiency tests developed with the Steering Committee for the Test of Proficiency-Huayu (SC-TOP). We evaluated various any-to-any models and vision-language models (VLMs) with audio transcription. Our results show that closed-source models generally outperform open-source ones across modalities, although open-source models can perform well in audio tasks. End-to-end any-to-any pipelines offer clear latency advantages compared to VLMs using separate audio transcription. Multi-TW presents a comprehensive view of model capabilities and highlights the need for Traditional Chinese fine-tuning and efficient multimodal architectures.

Large language models (LLMs) increasingly shape public understanding and civic decisions, yet their ideological neutrality is a growing concern. While existing research has explored various forms of LLM bias, a direct, cross-lingual comparison of models with differing geopolitical alignments-specifically a PRC-system model versus a non-PRC counterpart-has been lacking. This study addresses this gap by systematically evaluating DeepSeek-R1 (PRC-aligned) against ChatGPT o3-mini-high (non-PRC) for Chinese-state propaganda and anti-U.S. sentiment. We developed a novel corpus of 1,200 de-contextualized, reasoning-oriented questions derived from Chinese-language news, presented in Simplified Chinese, Traditional Chinese, and English. Answers from both models (7,200 total) were assessed using a hybrid evaluation pipeline combining rubric-guided GPT-4o scoring with human annotation. Our findings reveal significant model-level and language-dependent biases. DeepSeek-R1 consistently exhibited substantially higher proportions of both propaganda and anti-U.S. bias compared to ChatGPT o3-mini-high, which remained largely free of anti-U.S. sentiment and showed lower propaganda levels. For DeepSeek-R1, Simplified Chinese queries elicited the highest bias rates; these diminished in Traditional Chinese and were nearly absent in English. Notably, DeepSeek-R1 occasionally responded in Simplified Chinese to Traditional Chinese queries and amplified existing PRC-aligned terms in its Chinese answers, demonstrating an "invisible loudspeaker" effect. Furthermore, such biases were not confined to overtly political topics but also permeated cultural and lifestyle content, particularly in DeepSeek-R1.

While the capabilities of Large Language Models (LLMs) have been studied in both Simplified and Traditional Chinese, it is yet unclear whether LLMs exhibit differential performance when prompted in these two variants of written Chinese. This understanding is critical, as disparities in the quality of LLM responses can perpetuate representational harms by ignoring the different cultural contexts underlying Simplified versus Traditional Chinese, and can exacerbate downstream harms in LLM-facilitated decision-making in domains such as education or hiring. To investigate potential LLM performance disparities, we design two benchmark tasks that reflect real-world scenarios: regional term choice (prompting the LLM to name a described item which is referred to differently in Mainland China and Taiwan), and regional name choice (prompting the LLM to choose who to hire from a list of names in both Simplified and Traditional Chinese). For both tasks, we audit the performance of 11 leading commercial LLM services and open-sourced models -- spanning those primarily trained on English, Simplified Chinese, or Traditional Chinese. Our analyses indicate that biases in LLM responses are dependent on both the task and prompting language: while most LLMs disproportionately favored Simplified Chinese responses in the regional term choice task, they surprisingly favored Traditional Chinese names in the regional name choice task. We find that these disparities may arise from differences in training data representation, written character preferences, and tokenization of Simplified and Traditional Chinese. These findings highlight the need for further analysis of LLM biases; as such, we provide an open-sourced benchmark dataset to foster reproducible evaluations of future LLM behavior across Chinese language variants (https://github.com/brucelyu17/SC-TC-Bench).

In this paper, we propose a comprehensive evaluation benchmark for Visual Language Models (VLM) in Traditional Chinese. Our evaluation suite, the first of its kind, contains two complementary components: (1) VisTW-MCQ, a collection of manually curated exam multi-choice questions from 21 academic subjects designed to test the broad knowledge and reasoning capabilities of VLMs; and (2) VisTW-Dialogue, an open dialogue benchmark comprising 131 image-question pairs manually created to evaluate VLMs' ability in free-form dialogue generation within Taiwanese cultural contexts. These benchmarks address a critical gap in the evaluation landscape, where existing benchmarks predominantly focus on English or Simplified Chinese, neglecting the unique linguistic and cultural aspects of Traditional Chinese used in regions like Taiwan and Hong Kong. Our analysis reveals significant performance differences across various VLMs and highlights specific challenges in processing Traditional Chinese visual content.

Llama-Breeze2 (hereinafter referred to as Breeze2) is a suite of advanced multi-modal language models, available in 3B and 8B parameter configurations, specifically designed to enhance Traditional Chinese language representation. Building upon the Llama 3.2 model family, we continue the pre-training of Breeze2 on an extensive corpus to enhance the linguistic and cultural heritage of Traditional Chinese. In addition to language modeling capabilities, we significantly augment the models with function calling and vision understanding capabilities. At the time of this publication, as far as we are aware, absent reasoning-inducing prompts, Breeze2 are the strongest performing models in Traditional Chinese function calling and image understanding in its size class. The effectiveness of Breeze2 is benchmarked across various tasks, including Taiwan general knowledge, instruction-following, long context, function calling, and vision understanding. We are publicly releasing all Breeze2 models under the Llama 3.2 Community License. We also showcase the capabilities of the model running on mobile platform with a mobile application which we also open source.

The quality and size of a pretraining dataset significantly influence the performance of large language models (LLMs). While there have been numerous efforts in the curation of such a dataset for English users, there is a relative lack of similar initiatives for Traditional Chinese. Building upon this foundation of FineWeb, we introduce FineWeb-zhtw, a dataset tailored specifically for Traditional Chinese users. We came up with multiple stages of meticulously designed filters to cater to the linguistic difference between English and Traditional Chinese, to ensure comprehensiveness and quality. We determined effectiveness from querying dataset samples with three main objectives. Our code and datasets are publicly available.

We present TMMLU+, a new benchmark designed for Traditional Chinese language understanding. TMMLU+ is a multi-choice question-answering dataset with 66 subjects from elementary to professional level. It is six times larger and boasts a more balanced subject distribution than its predecessor, Taiwan Massive Multitask Language Understanding (TMMLU). We also benchmark closed-source models and 26 open-weight Chinese large language models (LLMs) of parameters ranging from 1.8B to 72B on the proposed TMMLU+. Our findings reveal that (1.) Traditional Chinese models still trail behind their Simplified Chinese counterparts, highlighting a need for more focused advancements in LLMs catering to Traditional Chinese. (2.) Current LLMs still fall short of human performance in average scores, indicating a potential need for future research to delve deeper into social science and humanities subjects. (3.) Among all the tokenization compression metrics examined, we identify that only the fertility score uniquely demonstrates strong correlations with our benchmark results. We foresee that TMMLU+ will pinpoint areas for future model improvement, thereby narrowing the gap between machine and human linguistic capabilities and supporting researchers in developing Traditional Chinese LLMs. Our dataset, along with the benchmark source code, is accessible at huggingface.co/datasets/ikala/tmmluplus.

The rapid development of large language models (LLMs) in recent years has largely focused on English, resulting in models that respond exclusively in English. To adapt these models to other languages, continual pre-training (CP) is often employed, followed by supervised fine-tuning (SFT) to maintain conversational abilities. However, CP and SFT can reduce a model's ability to filter harmful content. We propose Instruction Continual Pre-training (InsCP), which integrates instruction tags into the CP process to prevent loss of conversational proficiency while acquiring new languages. Our experiments demonstrate that InsCP retains conversational and Reinforcement Learning from Human Feedback (RLHF) abilities. Empirical evaluations on language alignment, reliability, and knowledge benchmarks confirm the efficacy of InsCP. Notably, this approach requires only 0.1 billion tokens of high-quality instruction-following data, thereby reducing resource consumption.

Large language models (LLMs) have demonstrated exceptional performance in various NLP applications. However, the majority of existing open-source LLMs are pre-trained primarily on English data and little part of other languages. This deficiency in multilingual training data results in suboptimal performance when applied to languages with fewer available resources. Furthermore, enhancing the performance of Large Language Models (LLMs) on low-resource languages by full-parameter fine-tuning with additional data requires substantial computational resources, posing computational barriers for research organizations and individual researchers. Consequently, several techniques such as parameter-efficient tuning and advanced embedding initialization have been proposed to address these challenges. In this work, we combine them to effectively facilitate cross-lingual transfer on English-dominated open-source LLM. Specifically, to effectively enhance the model's proficiency in Traditional Chinese, we conduct secondary pre-training on Llama 2 7B with Traditional Chinese data by leveraging QLoRA and our proposed zip-tie embedding initialization. The resulting model called Bailong, which stands for Bilingual trAnsfer learnIng based on qLOra and zip-tie embeddiNG. Recognizing the inadequacy of benchmark datasets in Traditional Chinese, we further introduce Bailong-bench to assess the alignment of models with human preferences and their capability to follow instructions in both Traditional Chinese and English tasks. In our evaluation, Bailong-instruct 7B exhibits competitive performance on Bailong-bench and other benchmark datasets when compared to other open-source models of similar or even larger parameter sizes. With rapid development in recent years, large language models (LLMs) have been widely utilized across various practical domains due to their remarkable abilities in contextual comprehension, reasoning, and text generation. One of the most notable applications is the use in chatbot implementations, which allows humans to interact directly with them through an intuitive user interface.