Yusuke Sakai

Riza Setiawan Soetedjo, Yusuke Sakai, Hidetaka Kamigaito et al.

Improving the quality of model-generated summaries, especially factuality, the accuracy of a summary with respect to its source content, remains a challenge. While reranking could select the optimal output from multiple generated candidates, it is limited to only using the source as guidance, resulting in unreliable summaries. To address this limitation, we propose ConSUM that reranks candidate summaries by considering two factors: consistency to the source document and consensus among the other candidates. Consensus is established using Minimum Bayes Risk (MBR) decoding over the set of generated summaries, while ensuring consistency by employing factuality-aware metrics that compare the summary against the source. Rigorous testing demonstrates that our system is competitive with existing methods, with human evaluations further confirming that its generated summaries are preferred over those from other systems. Our code is available at https://github.com/naist-nlp/ConSUM .

Haruki Sakajo, Frederikus Hudi, Yusuke Sakai et al.

Language exhibits inherent structures, a property that explains both language acquisition and language change. Given this characteristic, we expect language models to manifest their own internal structures as well. While interpretability research has investigated how models compute representations mechanistically through attention patterns and Sparse AutoEncoders, the organization of the resulting representations is overlooked. To address this gap, we introduce StructLens, a framework to analyze representations through a holistic structural view. StructLens constructs maximum spanning trees based on the semantic representations in residual streams, inspired by tree representation in dependency parsing, and provides summaries of token relationships in representation space. We analyze how contiguous tokens are also nearby in representation space and find that middle layers show the strongest local-span organization. Moreover, analysis of pre-training checkpoints reveals that smaller local units become detectable earlier in pre-training, and larger units later. Our findings demonstrate that StructLens provides insights into how models organize token representations across layers and training. Our code is available at https://github.com/naist-nlp/structlens.

Haruki Sakajo, Yusuke Sakai, Hidetaka Kamigaito et al.

Large language models (LLMs) have excelled in processing multiple languages through pre- and post-training on multilingual data, even though English dominates the training data. Prior work focusing on token representations has revealed how those LLMs process non-English text. Although these analyses have provided insightful findings, they fail to capture a structural view, which is an inherent property of language. In this study, we explore the multilinguality of LLMs through representational structural analysis. Our findings reveal that low-resource languages are structurally more different from English than high- and mid-resource languages, and that language-specific post-training alters their structures while preserving inter-language relationships.

Hiroyuki Deguchi, Yusuke Sakai, Hidetaka Kamigaito et al.

Minimum Bayes risk (MBR) decoding is a decision rule of text generation tasks that outperforms conventional maximum a posterior (MAP) decoding using beam search by selecting high-quality outputs based on a utility function rather than those with high-probability. Typically, it finds the most suitable hypothesis from the set of hypotheses under the sampled pseudo-references. mbrs is a library of MBR decoding, which can flexibly combine various metrics, alternative expectation estimations, and algorithmic variants. It is designed with a focus on speed measurement and calling count of code blocks, transparency, reproducibility, and extensibility, which are essential for researchers and developers. We published our mbrs as an MIT-licensed open-source project, and the code is available on GitHub. GitHub: https://github.com/naist-nlp/mbrs

Shintaro Ozaki, Kazuki Hayashi, Yusuke Sakai et al.

As the performance of Large-scale Vision Language Models (LVLMs) improves, they are increasingly capable of responding in multiple languages, and there is an expectation that the demand for explanations generated by LVLMs will grow. However, pre-training of Vision Encoder and the integrated training of LLMs with Vision Encoder are mainly conducted using English training data, leaving it uncertain whether LVLMs can completely handle their potential when generating explanations in languages other than English. In addition, multilingual QA benchmarks that create datasets using machine translation have cultural differences and biases, remaining issues for use as evaluation tasks. To address these challenges, this study created an extended dataset in multiple languages without relying on machine translation. This dataset that takes into account nuances and country-specific phrases was then used to evaluate the generation explanation abilities of LVLMs. Furthermore, this study examined whether Instruction-Tuning in resource-rich English improves performance in other languages. Our findings indicate that LVLMs perform worse in languages other than English compared to English. In addition, it was observed that LVLMs struggle to effectively manage the knowledge learned from English data. Our dataset is available at https://huggingface.co/datasets/naist-nlp/MultiExpArt

Chihiro Taguchi, Yusuke Sakai, Parisa Haghani et al.

This paper presents a state-of-the-art model for transcribing speech in any language into the International Phonetic Alphabet (IPA). Transcription of spoken languages into IPA is an essential yet time-consuming process in language documentation, and even partially automating this process has the potential to drastically speed up the documentation of endangered languages. Like the previous best speech-to-IPA model (Wav2Vec2Phoneme), our model is based on wav2vec 2.0 and is fine-tuned to predict IPA from audio input. We use training data from seven languages from CommonVoice 11.0, transcribed into IPA semi-automatically. Although this training dataset is much smaller than Wav2Vec2Phoneme's, its higher quality lets our model achieve comparable or better results. Furthermore, we show that the quality of our universal speech-to-IPA models is close to that of human annotators.

Takumi Goto, Yusuke Sakai, Taro Watanabe

Automatic evaluation in grammatical error correction (GEC) is crucial for selecting the best-performing systems. Currently, reference-based metrics are a popular choice, which basically measure the similarity between hypothesis and reference sentences. However, similarity measures based on embeddings, such as BERTScore, are often ineffective, since many words in the source sentences remain unchanged in both the hypothesis and the reference. This study focuses on edits specifically designed for GEC, i.e., ERRANT, and computes similarity measured over the edits from the source sentence. To this end, we propose edit vector, a representation for an edit, and introduce a new metric, UOT-ERRANT, which transports these edit vectors from hypothesis to reference using unbalanced optimal transport. Experiments with SEEDA meta-evaluation show that UOT-ERRANT improves evaluation performance, particularly in the +Fluency domain where many edits occur. Moreover, our method is highly interpretable because the transport plan can be interpreted as a soft edit alignment, making UOT-ERRANT a useful metric for both system ranking and analyzing GEC systems. Our code is available from https://github.com/gotutiyan/uot-errant.

Yusuke Ide, Yuto Nishida, Justin Vasselli et al.

The grammatical knowledge of language models (LMs) is often measured using a benchmark of linguistic minimal pairs, where the LMs are presented with a pair of acceptable and unacceptable sentences and required to judge which is more acceptable. Conventional approaches directly compare sentence probabilities assigned by LMs, but recent large language models (LLMs) are trained to perform tasks via prompting, and thus, the raw probabilities they assign may not fully reflect their grammatical knowledge. In this study, we attempt to derive more accurate acceptability judgments from LLMs using prompts and templates. Through extensive experiments in English and Chinese, we compare nine judgment methods and find two of them, a probability readout method -- in-template LP and a prompt-based method -- Yes/No probability computing, achieve higher accuracy than the conventional ones. Our analysis reveals that these methods excel in different linguistic phenomena, suggesting they access different aspects of LLMs' knowledge. We also find that ensembling the two methods outperforms single methods. Consequently, we recommend these techniques, either individually or ensembled, as more effective alternatives to conventional approaches for assessing grammatical knowledge in LLMs.

Yusuke Sakai, Adam Nohejl, Jiangnan Hang et al.

The natural language understanding (NLU) performance of large language models (LLMs) has been evaluated across various tasks and datasets. The existing evaluation methods, however, do not take into account the variance in scores due to differences in prompts, which leads to unfair evaluation and comparison of NLU performance. Moreover, evaluation designed for specific prompts is inappropriate for instruction tuning, which aims to perform well with any prompt. It is therefore necessary to find a way to measure NLU performance in a fair manner, considering score variance between different instruction templates. In this study, we provide English and Japanese cross-lingual datasets for evaluating the NLU performance of LLMs, which include multiple instruction templates for fair evaluation of each task, along with regular expressions to constrain the output format. Furthermore, we propose the Sharpe score as an evaluation metric that takes into account the variance in scores between templates. Comprehensive analysis of English and Japanese LLMs reveals that the high variance among templates has a significant impact on the fair evaluation of LLMs.

Yusuke Sakai, Hidetaka Kamigaito, Katsuhiko Hayashi et al.

Knowledge graphs (KGs) consist of links that describe relationships between entities. Due to the difficulty of manually enumerating all relationships between entities, automatically completing them is essential for KGs. Knowledge Graph Completion (KGC) is a task that infers unseen relationships between entities in a KG. Traditional embedding-based KGC methods, such as RESCAL, TransE, DistMult, ComplEx, RotatE, HAKE, HousE, etc., infer missing links using only the knowledge from training data. In contrast, the recent Pre-trained Language Model (PLM)-based KGC utilizes knowledge obtained during pre-training. Therefore, PLM-based KGC can estimate missing links between entities by reusing memorized knowledge from pre-training without inference. This approach is problematic because building KGC models aims to infer unseen links between entities. However, conventional evaluations in KGC do not consider inference and memorization abilities separately. Thus, a PLM-based KGC method, which achieves high performance in current KGC evaluations, may be ineffective in practical applications. To address this issue, we analyze whether PLM-based KGC methods make inferences or merely access memorized knowledge. For this purpose, we propose a method for constructing synthetic datasets specified in this analysis and conclude that PLMs acquire the inference abilities required for KGC through pre-training, even though the performance improvements mostly come from textual information of entities and relations.

Yusuke Sakai, Hidetaka Kamigaito, Taro Watanabe

Recently, we have often observed hallucinated citations or references that do not correspond to any existing work in papers under review, preprints, or published papers. Such hallucinated citations pose a serious concern to scientific reliability. When they appear in accepted papers, they may also negatively affect the credibility of conferences. In this study, we refer to hallucinated citations as "HalluCitation" and systematically investigate their prevalence and impact. We analyze all papers published at ACL, NAACL, and EMNLP in 2024 and 2025, including main conference, Findings, and workshop papers. Our analysis reveals that nearly 300 papers contain at least one HalluCitation, most of which were published in 2025. Notably, half of these papers were identified at EMNLP 2025, the most recent conference, indicating that this issue is rapidly increasing. Moreover, more than 100 such papers were accepted as main conference and Findings papers at EMNLP 2025, affecting the credibility.

Xuefeng Wei, Zhixuan Wang, Xuan Zhou et al.

We introduce CARTBENCH, a museum-grounded benchmark for evaluating vision-language models (VLMs) on Chinese artworks beyond short-form recognition and QA. CARTBENCH comprises four subtasks: CURATORQA for evidence-grounded recognition and reasoning, CATALOGCAPTION for structured four-section expert-style appreciation, REINTERPRET for defensible reinterpretation with expert ratings, and CONNOISSEURPAIRS for diagnostic authenticity discrimination under visually similar confounds. CARTBENCH is built by aligning image-bearing Palace Museum objects from Wikidata with authoritative catalog pages, spanning five art categories across multiple dynasties. Across nine representative VLMs, we find that high overall CURATORQA accuracy can mask sharp drops on hard evidence linking and style-to-period inference; long-form appreciation remains far from expert references; and authenticity-oriented diagnostic discrimination stays near chance, underscoring the difficulty of connoisseur-level reasoning for current models.

Peinan Zhang, Yusuke Sakai, Masato Mita et al.

With the increase in the fluency of ad texts automatically created by natural language generation technology, there is high demand to verify the quality of these creatives in a real-world setting. We propose AdTEC (Ad Text Evaluation Benchmark by CyberAgent), the first public benchmark to evaluate ad texts from multiple perspectives within practical advertising operations. Our contributions are as follows: (i) Defining five tasks for evaluating the quality of ad texts, as well as building a Japanese dataset based on the practical operational experiences of building a Japanese dataset based on the practical operational experiences of advertising agencies, which are typically kept in-house. (ii) Validating the performance of existing pre-trained language models (PLMs) and human evaluators on the dataset. (iii) Analyzing the characteristics and providing challenges of the benchmark. The results show that while PLMs have already reached practical usage level in several tasks, humans still outperform in certain domains, implying that there is significant room for improvement in this area.

Yuto Omae, Kazuki Sakai, Yohei Kakimoto et al.

Neural networks (NNs) are central to modern machine learning and achieve state-of-the-art results in many applications. However, the relationship between loss geometry and generalization is still not well understood. The local geometry of the loss function near a critical point is well-approximated by its quadratic form, obtained through a second-order Taylor expansion. The coefficients of the quadratic term correspond to the Hessian matrix, whose eigenspectrum allows us to evaluate the sharpness of the loss at the critical point. Extensive research suggests flat critical points generalize better, while sharp ones lead to higher generalization error. However, sharpness requires the Hessian eigenspectrum, but general matrix characteristic equations have no closed-form solution. Therefore, most existing studies on evaluating loss sharpness rely on numerical approximation methods. Existing closed-form analyses of the eigenspectrum are primarily limited to simplified architectures, such as linear or ReLU-activated networks; consequently, theoretical analysis of smooth nonlinear multilayer neural networks remains limited. Against this background, this study focuses on nonlinear, smooth multilayer neural networks and derives a closed-form upper bound for the maximum eigenvalue of the Hessian with respect to the cross-entropy loss by leveraging the Wolkowicz-Styan bound. Specifically, the derived upper bound is expressed as a function of the affine transformation parameters, hidden layer dimensions, and the degree of orthogonality among the training samples. The primary contribution of this paper is an analytical characterization of loss sharpness in smooth nonlinear multilayer neural networks via a closed-form expression, avoiding explicit numerical eigenspectrum computation. We hope that this work provides a small yet meaningful step toward unraveling the mysteries of deep learning.

Yusuke Sakai, Mana Makinae, Hidetaka Kamigaito et al.

In Simultaneous Machine Translation (SiMT) systems, training with a simultaneous interpretation (SI) corpus is an effective method for achieving high-quality yet low-latency systems. However, it is very challenging to curate such a corpus due to limitations in the abilities of annotators, and hence, existing SI corpora are limited. Therefore, we propose a method to convert existing speech translation corpora into interpretation-style data, maintaining the original word order and preserving the entire source content using Large Language Models (LLM-SI-Corpus). We demonstrate that fine-tuning SiMT models in text-to-text and speech-to-text settings with the LLM-SI-Corpus reduces latencies while maintaining the same level of quality as the models trained with offline datasets. The LLM-SI-Corpus is available at \url{https://github.com/yusuke1997/LLM-SI-Corpus}.

Yuto Omae, Yusuke Sakai, Hirotaka Takahashi

Recently, many convolutional neural networks (CNNs) for classification by time domain data of multisignals have been developed. Although some signals are important for correct classification, others are not. When data that do not include important signals for classification are taken as the CNN input layer, the calculation, memory, and data collection costs increase. Therefore, identifying and eliminating nonimportant signals from the input layer are important. In this study, we proposed features gradient-based signals selection algorithm (FG-SSA), which can be used for finding and removing nonimportant signals for classification by utilizing features gradient obtained by the calculation process of grad-CAM. When we define N as the number of signals, the computational complexity of the proposed algorithm is linear time O(N), that is, it has a low calculation cost. We verified the effectiveness of the algorithm using the OPPORTUNITY Activity Recognition dataset, which is an open dataset comprising acceleration signals of human activities. In addition, we checked the average 6.55 signals from a total of 15 acceleration signals (five triaxial sensors) that were removed by FG-SSA while maintaining high generalization scores of classification. Therefore, the proposed algorithm FG-SSA has an effect on finding and removing signals that are not important for CNN-based classification.

Xincan Feng, Noriki Nishida, Yusuke Sakai et al.

Prior work reports conflicting results on query diversity in synthetic data generation for dense retrieval. We identify this conflict and design Q-D metrics to quantify diversity's impact, making the problem measurable. Through experiments on 4 benchmark types (31 datasets), we find query diversity especially benefits multi-hop retrieval. Deep analysis on multi-hop data reveals that diversity benefit correlates strongly with query complexity ($r$$\geq$0.95, $p$$<$0.05 in 12/14 conditions), measured by content words (CW). We formalize this as the Complexity-Diversity Principle (CDP): query complexity determines optimal diversity. CDP provides actionable thresholds (CW$>$10: use diversity; CW$<$7: avoid it). Guided by CDP, we propose zero-shot multi-query synthesis for multi-hop tasks, achieving state-of-the-art performance.

Takumi Goto, Yusuke Sakai, Taro Watanabe

Grammatical error correction using large language models often suffers from the over-correction issue. To mitigate this, we propose a training-free inference method that performs edit-level majority voting over multiple candidates generated by a single model, without requiring model modifications or additional training. Across nine benchmarks covering English, Czech, German, Ukrainian, Korean, Hindi, and Romanian, the proposed method outperforms both greedy and MBR decoding in most cases. Moreover, it yields stable correction quality regardless of the instruction prompts used. We release two repository supporting GEC datasets loading and LLM inference.

Koki Natsumi, Hiroyuki Deguchi, Yusuke Sakai et al.

Minimum Bayes risk (MBR) decoding generates high-quality translations by maximizing the expected utility of output candidates, but it evaluates all pairwise scores over the candidate set; hence, it takes quadratic time with respect to the number of candidates. To reduce the number of utility function calls, probabilistic MBR (PMBR) decoding partially evaluates quality scores using sampled pairs of candidates and completes the missing scores with a matrix completion algorithm. Nevertheless, it degrades the translation quality as the number of utility function calls is reduced. Therefore, to improve the trade-off between quality and cost, we propose agreement-constrained PMBR (AC-PMBR) decoding, which leverages a knowledge distilled model to guide the completion of the score matrix. Our AC-PMBR decoding improved approximation errors of matrix completion by up to 3 times and achieved higher translation quality compared with PMBR decoding at a comparable computational cost on the WMT'23 En$\leftrightarrow$De translation tasks.

Hiroyuki Deguchi, Katsuki Chousa, Yusuke Sakai

Strongly human-correlated evaluation metrics serve as an essential compass for the development and improvement of generation models and must be highly reliable and robust. Recent embedding-based neural text evaluation metrics, such as COMET for translation tasks, are widely used in both research and development fields. However, there is no guarantee that they yield reliable evaluation results due to the black-box nature of neural networks. To raise concerns about the reliability and safety of such metrics, we propose a method for finding a single adversarial text in the discrete space that is consistently evaluated as high-quality, regardless of the test cases, to identify the vulnerabilities in evaluation metrics. The single hub text found with our method achieved 79.1 COMET% and 67.8 COMET% in the WMT'24 English-to-Japanese (En--Ja) and English-to-German (En--De) translation tasks, respectively, outperforming translations generated individually for each source sentence by using M2M100, a general translation model. Furthermore, we also confirmed that the hub text found with our method generalizes across multiple language pairs such as Ja--En and De--En.

Kazuki Hayashi, Yusuke Sakai, Hidetaka Kamigaito et al.

Large-scale vision-language models (LVLMs) output text from images and instructions, demonstrating advanced capabilities in text generation and comprehension. However, it has not been clarified to what extent LVLMs understand the knowledge necessary for explaining images, the complex relationships between various pieces of knowledge, and how they integrate these understandings into their explanations. To address this issue, we propose a new task: the artwork explanation generation task, along with its evaluation dataset and metric for quantitatively assessing the understanding and utilization of knowledge about artworks. This task is apt for image description based on the premise that LVLMs are expected to have pre-existing knowledge of artworks, which are often subjects of wide recognition and documented information. It consists of two parts: generating explanations from both images and titles of artworks, and generating explanations using only images, thus evaluating the LVLMs' language-based and vision-based knowledge. Alongside, we release a training dataset for LVLMs to learn explanations that incorporate knowledge about artworks. Our findings indicate that LVLMs not only struggle with integrating language and visual information but also exhibit a more pronounced limitation in acquiring knowledge from images alone. The datasets (ExpArt=Explain Artworks) are available at https://huggingface.co/datasets/naist-nlp/ExpArt.

Haruki Sakajo, Yusuke Sakai, Hidetaka Kamigaito et al.

Vowels are primarily characterized by tongue position. Humans have discovered these features of vowel articulation through their own experience and explicit objective observation such as using MRI. With this knowledge and our experience, we can explain and understand the relationship between tongue positions and vowels, and this knowledge is helpful for language learners to learn pronunciation. Since language models (LMs) are trained on a large amount of data that includes linguistic and medical fields, our preliminary studies indicate that an LM is able to explain the pronunciation mechanisms of vowels. However, it is unclear whether multi-modal LMs, such as vision LMs, align textual information with visual information. One question arises: do LMs associate real tongue positions with vowel articulation? In this study, we created video and image datasets from the existing real-time MRI dataset and investigated whether LMs can understand vowel articulation based on tongue positions using vision-based information. Our findings suggest that LMs exhibit potential for understanding vowels and tongue positions when reference examples are provided while they have difficulties without them. Our code for dataset building is available on GitHub.

Hidetaka Kamigaito, Hiroyuki Deguchi, Yusuke Sakai et al.

Inference methods play an important role in eliciting the performance of large language models (LLMs). Currently, LLMs use inference methods utilizing generated multiple samples, which can be derived from Minimum Bayes Risk (MBR) Decoding. Previous studies have conducted empirical analyses to clarify the improvements in generation performance achieved by MBR decoding and have reported various observations. However, the theoretical underpinnings of these findings remain uncertain. To address this, we offer a new theoretical interpretation of MBR decoding from the perspective of bias-diversity decomposition. In this interpretation, the error in the quality estimation of hypotheses by MBR decoding is decomposed into two main factors: bias, which considers the closeness between the utility function and human evaluation, and diversity, which represents the variability in the quality estimation of the utility function. The theoretical analysis reveals the difficulty of simultaneously improving bias and diversity, confirming the validity of enhancing MBR decoding performance by increasing diversity. Furthermore, we reveal that diversity can explain one aspect of inference scaling laws that describe performance improvement by increasing sample size. Moreover, experiments across multiple NLP tasks yielded results consistent with these theoretical characteristics. Our code is available at https://github.com/naist-nlp/mbr-bias-diversity.

Takumi Goto, Yusuke Sakai, Taro Watanabe

We introduce gec-metrics, a library for using and developing grammatical error correction (GEC) evaluation metrics through a unified interface. Our library enables fair system comparisons by ensuring that everyone conducts evaluations using a consistent implementation. Moreover, it is designed with a strong focus on API usage, making it highly extensible. It also includes meta-evaluation functionalities and provides analysis and visualization scripts, contributing to developing GEC evaluation metrics. Our code is released under the MIT license and is also distributed as an installable package. The video is available on YouTube.

Yusuke Sakai, Hidetaka Kamigaito, Taro Watanabe

It is very challenging to curate a dataset for language-specific knowledge and common sense in order to evaluate natural language understanding capabilities of language models. Due to the limitation in the availability of annotators, most current multilingual datasets are created through translation, which cannot evaluate such language-specific aspects. Therefore, we propose Multilingual CommonsenseQA (mCSQA) based on the construction process of CSQA but leveraging language models for a more efficient construction, e.g., by asking LM to generate questions/answers, refine answers and verify QAs followed by reduced human efforts for verification. Constructed dataset is a benchmark for cross-lingual language-transfer capabilities of multilingual LMs, and experimental results showed high language-transfer capabilities for questions that LMs could easily solve, but lower transfer capabilities for questions requiring deep knowledge or commonsense. This highlights the necessity of language-specific datasets for evaluation and training. Finally, our method demonstrated that multilingual LMs could create QA including language-specific knowledge, significantly reducing the dataset creation cost compared to manual creation. The datasets are available at https://huggingface.co/datasets/yusuke1997/mCSQA.

Hiroyuki Deguchi, Katsuki Chousa, Yusuke Sakai

The hubness problem, in which hub embeddings are close to many unrelated examples, occurs often in high-dimensional embedding spaces and may pose a practical threat for purposes such as information retrieval and automatic evaluation metrics. In particular, since cross-modal similarity between text and images cannot be calculated by direct comparisons, such as string matching, cross-modal encoders that project different modalities into a shared space are helpful for various cross-modal applications, and thus, the existence of hubs may pose practical threats. To reveal the vulnerabilities of cross-modal encoders, we propose a method for identifying the hub embedding and its corresponding hub text. Experiments on image captioning evaluation in MSCOCO and nocaps along with image-to-text retrieval tasks in MSCOCO and Flickr30k showed that our method can identify a single hub text that unreasonably achieves comparable or higher similarity scores than human-written reference captions in many images, thereby revealing the vulnerabilities in cross-modal encoders.

Yusuke Sakai, Hidetaka Kamigaito, Taro Watanabe

We introduce HalluCiteChecker, a toolkit for detecting and verifying hallucinated citations in scientific papers. While AI assistant technologies have transformed the academic writing process, including citation recommendation, they have also led to the emergence of hallucinated citations that do not correspond to any existing work. Such citations not only undermine the credibility of scientific papers but also impose an additional burden on reviewers and authors, who must manually verify their validity during the review process. In this study, we formalize hallucinated citation detection as an NLP task and provide a corresponding toolkit as a practical foundation for addressing this problem. Our package is lightweight and can perform verification in seconds on a standard laptop. It can also be executed entirely offline and runs efficiently using only CPUs. We hope that HalluCiteChecker will help reduce reviewer workload and support organizers by enabling systematic pre-review and publication checks. Our code is released under the Apache 2.0 license on GitHub and is distributed as an installable package via PyPI. A demonstration video is available on YouTube.

Hiroyuki Deguchi, Yusuke Sakai, Hidetaka Kamigaito et al.

Minimum Bayes risk (MBR) decoding achieved state-of-the-art translation performance by using COMET, a neural metric that has a high correlation with human evaluation. However, MBR decoding requires quadratic time since it computes the expected score between a translation hypothesis and all reference translations. We propose centroid-based MBR (CBMBR) decoding to improve the speed of MBR decoding. Our method clusters the reference translations in the feature space, and then calculates the score using the centroids of each cluster. The experimental results show that our CBMBR not only improved the decoding speed of the expected score calculation 5.7 times, but also outperformed vanilla MBR decoding in translation quality by up to 0.5 COMET in the WMT'22 En$\leftrightarrow$Ja, En$\leftrightarrow$De, En$\leftrightarrow$Zh, and WMT'23 En$\leftrightarrow$Ja translation tasks.

Takumi Goto, Yusuke Sakai, Taro Watanabe

One of the goals of automatic evaluation metrics in grammatical error correction (GEC) is to rank GEC systems such that it matches human preferences. However, current automatic evaluations are based on procedures that diverge from human evaluation. Specifically, human evaluation derives rankings by aggregating sentence-level relative evaluation results, e.g., pairwise comparisons, using a rating algorithm, whereas automatic evaluation averages sentence-level absolute scores to obtain corpus-level scores, which are then sorted to determine rankings. In this study, we propose an aggregation method for existing automatic evaluation metrics which aligns with human evaluation methods to bridge this gap. We conducted experiments using various metrics, including edit-based metrics, n-gram based metrics, and sentence-level metrics, and show that resolving the gap improves results for the most of metrics on the SEEDA benchmark. We also found that even BERT-based metrics sometimes outperform the metrics of GPT-4. The proposed ranking method is integrated gec-metrics.

Shintaro Ozaki, Kazuki Hayashi, Miyu Oba et al.

A large part of human communication relies on nonverbal cues such as facial expressions, eye contact, and body language. Unlike language or sign language, such nonverbal communication lacks formal rules, requiring complex reasoning based on commonsense understanding. Enabling current Video Large Language Models (VideoLLMs) to accurately interpret body language is a crucial challenge, as human unconscious actions can easily cause the model to misinterpret their intent. To address this, we propose a dataset, BQA, a body language question answering dataset, to validate whether the model can correctly interpret emotions from short clips of body language comprising 26 emotion labels of videos of body language. We evaluated various VideoLLMs on BQA and revealed that understanding body language is challenging, and our analyses of the wrong answers by VideoLLMs show that certain VideoLLMs made significantly biased answers depending on the age group and ethnicity of the individuals in the video. The dataset is available.

Yusuke Sakai, Hidetaka Kamigaito, Taro Watanabe

In generative commonsense reasoning tasks such as CommonGen, generative large language models (LLMs) compose sentences that include all given concepts. However, when focusing on instruction-following capabilities, if a prompt specifies a concept order, LLMs must generate sentences that adhere to the specified order. To address this, we propose Ordered CommonGen, a benchmark designed to evaluate the compositional generalization and instruction-following abilities of LLMs. This benchmark measures ordered coverage to assess whether concepts are generated in the specified order, enabling a simultaneous evaluation of both abilities. We conducted a comprehensive analysis using 36 LLMs and found that, while LLMs generally understand the intent of instructions, biases toward specific concept order patterns often lead to low-diversity outputs or identical results even when the concept order is altered. Moreover, even the most instruction-compliant LLM achieved only about 75% ordered coverage, highlighting the need for improvements in both instruction-following and compositional generalization capabilities.

Yusuke Sakai, Takumi Goto, Taro Watanabe

We propose IMPARA-GED, a novel reference-free automatic grammatical error correction (GEC) evaluation method with grammatical error detection (GED) capabilities. We focus on the quality estimator of IMPARA, an existing automatic GEC evaluation method, and construct that of IMPARA-GED using a pre-trained language model with enhanced GED capabilities. Experimental results on SEEDA, a meta-evaluation dataset for automatic GEC evaluation methods, demonstrate that IMPARA-GED achieves the highest correlation with human sentence-level evaluations.

Kazuki Hayashi, Shintaro Ozaki, Yusuke Sakai et al.

Large-scale Vision Language Models (LVLMs) are increasingly being applied to a wide range of real-world multimodal applications, involving complex visual and linguistic reasoning. As these models become more integrated into practical use, they are expected to handle complex aspects of human interaction. Among these, color perception is a fundamental yet highly variable aspect of visual understanding. It differs across individuals due to biological factors such as Color Vision Deficiencies (CVDs), as well as differences in culture and language. Despite its importance, perceptual diversity has received limited attention. In our study, we evaluate LVLMs' ability to account for individual level perceptual variation using the Ishihara Test, a widely used method for detecting CVDs. Our results show that LVLMs can explain CVDs in natural language, but they cannot simulate how people with CVDs perceive color in image based tasks. These findings highlight the need for multimodal systems that can account for color perceptual diversity and support broader discussions on perceptual inclusiveness and fairness in multimodal AI.

Takumi Goto, Yusuke Sakai, Taro Watanabe

Reference-free evaluation metrics for grammatical error correction (GEC) have achieved high correlation with human judgments. However, these metrics are not designed to evaluate adversarial systems that aim to obtain unjustifiably high scores. The existence of such systems undermines the reliability of automatic evaluation, as it can mislead users in selecting appropriate GEC systems. In this study, we propose adversarial attack strategies for four reference-free metrics: SOME, Scribendi, IMPARA, and LLM-based metrics, and demonstrate that our adversarial systems outperform the current state-of-the-art. These findings highlight the need for more robust evaluation methods.

Justin Vasselli, Eunike Andriani Kardinata, Yusuke Sakai et al.

Non-English dialogue datasets are scarce, and models are often trained or evaluated on translations of English-language dialogues, an approach which can introduce artifacts that reduce their naturalness and cultural appropriateness. This work proposes Dialogue Act Script (DAS), a structured framework for encoding, localizing, and generating multilingual dialogues from abstract intent representations. Rather than translating dialogue utterances directly, DAS enables the generation of new dialogues in the target language that are culturally and contextually appropriate. By using structured dialogue act representations, DAS supports flexible localization across languages, mitigating translationese and enabling more fluent, naturalistic conversations. Human evaluations across Italian, German, and Chinese show that DAS-generated dialogues consistently outperform those produced by both machine and human translators on measures of cultural relevance, coherence, and situational appropriateness.

Ashmari Pramodya, Nirasha Nelki, Heshan Shalinda et al.

Large Language Models (LLMs) demonstrate impressive general knowledge and reasoning abilities, yet their evaluation has predominantly focused on global or anglocentric subjects, often neglecting low-resource languages and culturally specific content. While recent multilingual benchmarks attempt to bridge this gap, many rely on automatic translation, which can introduce errors and misrepresent the original cultural context. To address this, we introduce SinhalaMMLU, the first multiple-choice question answering benchmark designed specifically for Sinhala, a low-resource language. The dataset includes over 7,000 questions spanning secondary to collegiate education levels, aligned with the Sri Lankan national curriculum, and covers six domains and 30 subjects, encompassing both general academic topics and culturally grounded knowledge. We evaluate 26 LLMs on SinhalaMMLU and observe that, while Claude 3.5 sonnet and GPT-4o achieve the highest average accuracies at 67% and 62% respectively, overall model performance remains limited. In particular, models struggle in culturally rich domains such as the Humanities, revealing substantial room for improvement in adapting LLMs to low-resource and culturally specific contexts.

Chihiro Taguchi, Seng Mai, Keita Kurabe et al.

Multilingual machine translation (MT) benchmarks play a central role in evaluating the capabilities of modern MT systems. Among them, the FLORES+ benchmark is widely used, offering English-to-many translation data for over 200 languages, curated with strict quality control protocols. However, we study data in four languages (Asante Twi, Japanese, Jinghpaw, and South Azerbaijani) and uncover critical shortcomings in the benchmark's suitability for truly multilingual evaluation. Human assessments reveal that many translations fall below the claimed 90% quality standard, and the annotators report that source sentences are often too domain-specific and culturally biased toward the English-speaking world. We further demonstrate that simple heuristics, such as copying named entities, can yield non-trivial BLEU scores, suggesting vulnerabilities in the evaluation protocol. Notably, we show that MT models trained on high-quality, naturalistic data perform poorly on FLORES+ while achieving significant gains on our domain-relevant evaluation set. Based on these findings, we advocate for multilingual MT benchmarks that use domain-general and culturally neutral source texts rely less on named entities, in order to better reflect real-world translation challenges.

Haruki Sakajo, Yusuke Ide, Justin Vasselli et al.

Cross-lingual vocabulary transfer plays a promising role in adapting pre-trained language models to new languages, including low-resource languages. Existing approaches that utilize monolingual or parallel corpora face challenges when applied to languages with limited resources. In this work, we propose a simple yet effective vocabulary transfer method that utilizes bilingual dictionaries, which are available for many languages, thanks to descriptive linguists. Our proposed method leverages a property of BPE tokenizers where removing a subword from the vocabulary causes a fallback to shorter subwords. The embeddings of target subwords are estimated iteratively by progressively removing them from the tokenizer. The experimental results show that our approach outperforms existing methods for low-resource languages, demonstrating the effectiveness of a dictionary-based approach for cross-lingual vocabulary transfer.

Shintaro Ozaki, Kazuki Hayashi, Yusuke Sakai et al.

Generating images from prompts containing specific entities requires models to retain as much entity-specific knowledge as possible. However, fully memorizing such knowledge is impractical due to the vast number of entities and their continuous emergence. To address this, we propose Text-based Intelligent Generation with Entity prompt Refinement (TextTIGER), which augments knowledge on entities included in the prompts and then summarizes the augmented descriptions using Large Language Models (LLMs) to mitigate performance degradation from longer inputs. To evaluate our method, we introduce WiT-Cub (WiT with Captions and Uncomplicated Background-explanations), a dataset comprising captions, images, and an entity list. Experiments on four image generation models and five LLMs show that TextTIGER improves image generation performance in standard metrics (IS, FID, and CLIPScore) compared to caption-only prompts. Additionally, multiple annotators' evaluation confirms that the summarized descriptions are more informative, validating LLMs' ability to generate concise yet rich descriptions. These findings demonstrate that refining prompts with augmented and summarized entity-related descriptions enhances image generation capabilities. The code and dataset will be available upon acceptance.

Takashi Harada, Takehiro Motomitsu, Katsuhiko Hayashi et al.

In recent years, there has been a notable increase in research on machine learning models for music retrieval and generation systems that are capable of taking natural language sentences as inputs. However, there is a scarcity of large-scale publicly available datasets, consisting of music data and their corresponding natural language descriptions known as music captions. In particular, non-musical information such as suitable situations for listening to a track and the emotions elicited upon listening is crucial for describing music. This type of information is underrepresented in existing music caption datasets due to the challenges associated with extracting it directly from music data. To address this issue, we propose a method for generating music caption data that incorporates non-musical aspects inferred from music thumbnail images, and validated the effectiveness of our approach through human evaluations. Additionally, we created a dataset with approximately 360,000 captions containing non-musical aspects. Leveraging this dataset, we trained a music retrieval model and demonstrated its effectiveness in music retrieval tasks through evaluation.

Kazuki Hayashi, Kazuma Onishi, Toma Suzuki et al.

Large-scale Vision-Language Models (LVLMs) process both images and text, excelling in multimodal tasks such as image captioning and description generation. However, while these models excel at generating factual content, their ability to generate and evaluate texts reflecting perspectives on the same image, depending on the context, has not been sufficiently explored. To address this, we propose IRR: Image Review Rank, a novel evaluation framework designed to assess critic review texts from multiple perspectives. IRR evaluates LVLMs by measuring how closely their judgments align with human interpretations. We validate it using a dataset of images from 15 categories, each with five critic review texts and annotated rankings in both English and Japanese, totaling over 2,000 data instances. The datasets are available at https://hf.co/datasets/naist-nlp/Wiki-ImageReview1.0. Our results indicate that, although LVLMs exhibited consistent performance across languages, their correlation with human annotations was insufficient, highlighting the need for further advancements. These findings highlight the limitations of current evaluation methods and the need for approaches that better capture human reasoning in Vision & Language tasks.

Yusuke Sakai, Yousuke Itoh, Piljong Jung et al.

In the data obtained by laser interferometric gravitational wave detectors, transient noise with non-stationary and non-Gaussian features occurs at a high rate. This often results in problems such as detector instability and the hiding and/or imitation of gravitational-wave signals. This transient noise has various characteristics in the time--frequency representation, which is considered to be associated with environmental and instrumental origins. Classification of transient noise can offer clues for exploring its origin and improving the performance of the detector. One approach for accomplishing this is supervised learning. However, in general, supervised learning requires annotation of the training data, and there are issues with ensuring objectivity in the classification and its corresponding new classes. By contrast, unsupervised learning can reduce the annotation work for the training data and ensure objectivity in the classification and its corresponding new classes. In this study, we propose an unsupervised learning architecture for the classification of transient noise that combines a variational autoencoder and invariant information clustering. To evaluate the effectiveness of the proposed architecture, we used the dataset (time--frequency two-dimensional spectrogram images and labels) of the Laser Interferometer Gravitational-wave Observatory (LIGO) first observation run prepared by the Gravity Spy project. The classes provided by our proposed unsupervised learning architecture were consistent with the labels annotated by the Gravity Spy project, which manifests the potential for the existence of unrevealed classes.