arXiv subjects

Ondřej Bojar

Publications and source records attributed to Ondřej Bojar.

At least 19 recordsLinked to original sources

arXiv2026

Current state of LLMs for Arabic dialectal machine translation

This work presents an evaluation of large language models (LLMs) for English to dialectal Arabic machine translation on the MADAR dataset. We evaluate both translation directions (English to Arabic and vice-versa) on 16 Arabic dialects. Our experiments cover a diverse set of models, including specialized Arabic models (Jais, Nile), multilingual models (Gemma, Command-R, Mistral, Aya), and commercial APIs (GPT-4.1). We employ multiple evaluation metrics: BLEU, CHRF, COMET (both reference-based and reference-less variants) and GEMBA (LLM-as-a-judge), as well as a small-scale manual evaluation, to assess translation quality. We discuss the challenges of automatic MT evaluation, especially in the context of Arabic dialects. We also evaluate the ability of LLMs to classify the dialect used in a text. The study offers insights into the capabilities and limitations of current LLMs for dialectal Arabic machine translation, particularly highlighting the difficulty of handling dialectal diversity, although the results may be influenced by possible training data contamination, which is always a concern with LLMs.

cs.CL↗

arXiv2026

ImgMT: Wikipedia-based Dataset For Evaluating Text-in-Image Translation

Text-in-image translation aims to translate all textual content embedded in an image. In Text Image Translation (TIT), the output is translated text; In-Image Machine Translation (IIMT) additionally renders the translation back into the image. We introduce ImgMT, a massively multilingual dataset and benchmark mined from localized SVG illustrations on Wikimedia Commons, focusing on diagrams, maps, and infographics. ImgMT contains 35,612 aligned image pairs grouped into 1,855 image sets, spanning 129 languages, 20+ writing systems, and 4,886 language pairs. We propose a multi-level evaluation protocol that measures source-side text-region detection and OCR accuracy alongside target-side translation quality. We benchmark several multimodal large language models and a dedicated TIT system in English->X and X->English settings, and find that the source language and writing system substantially affect TIT quality. Finally, we show that ImgMT is not well suited for isolating the contribution of visual information in multimodal translation.

cs.CL↗

arXiv2026

Last Translation Benchmark

For scientific progress, we need benchmarks that test the limits of state-of-the-art models, and evaluation methods that inform us about failure cases. As models get stronger, standard benchmarks for machine translation are approaching saturation. Further, automatic translation metrics are unreliable, opaque, and vulnerable to reward-hacking. Even gold human evaluation is not problem-free, because it often lacks reproducibility, objectivity, and scalability. Overall, this prevents us from tracking progress in the field and identifying pathways for improvement. We introduce the Last Translation Benchmark, a collection of human-authored and peer-reviewed examples (texts, images, audio, videos) that break leading machine translation models. We also present a new evaluation approach: each example comes with handcrafted verification rules describing concrete failure cases on that example, therefore allowing reliable and actionable future evaluation. The Last Translation Benchmark is a live dataset that accepts ongoing contributions. The latest version is LTBv1, containing accepted contributions prior to September 1st 2026, with future releases planned as new data is continuously collected.

cs.CL↗

arXiv2026

Dynamically Allocating Evaluation Effort for Model Ranking

While human evaluation is the gold standard in many NLP tasks, it suffers from prohibitive costs and poor scalability. When identifying top-performing models, typical evaluation protocols waste effort by exhaustively evaluating all models on the entire benchmark, a safe but inefficient approach. In this work, we formalize multi-model human evaluation as a best-arm identification problem in a multi-armed bandit setup with correlated arms, where pulling an arm corresponds to human-evaluating a model. By sampling adaptively based on the intermediate model rankings obtained on the samples so far, we can focus the annotation budget on the most competitive models. We prove the optimality of the proposed algorithms and show that it improves discrimination between top-performing models. This makes evaluations faster, cheaper and more aligned with large-scale competition evaluation goals.

cs.CL↗

arXiv2026

Instability of LLM Pre-Pretraining: It Doesn't Always Help. An Investigation on Multiple Languages

Pretraining LLMs on artificial languages ("pre-pretraining") is a technique that could reportedly increase token efficiency by 33%, i.e., save up to 33% of training tokens needed to reach a certain performance. We validate this prior result for English on a larger set of natural languages across four language families, using two different tokenizers and varying model sizes. We also relate the observed gains (or losses) in token efficiency to quantified linguistic properties of the languages, such as sentence length, morphological richness, and features of dependency syntactic trees (tree depth, number of children, number of crossing dependencies). Our empirical results indicate that the reported gains depend heavily on the experiment setup and the choice of random seed, although we can confirm the trend of stable gains with 128-Dyck pretraining of small models with the Llama tokenizer for most of the examined languages. On a general note, we argue that multiple training runs should be carried out at least for a subset of experiments to avoid the community adopting unstable approaches.

cs.CL↗

arXiv2026

CzechDocs: A Multiway Parallel Dataset of Formatted Documents for Minority Languages in Czechia

We present CzechDocs, a multiway parallel dataset of formatted documents (HTML, DOCX, and PDF) covering Czech and minority languages used in Czechia-primarily Ukrainian and English, with smaller portions of Vietnamese, Russian and other languages. The dataset is designed to support the evaluation of machine translation systems that aim to preserve document formatting during translation. We provide a comparison of the most common approaches to format-preserving machine translation on a validation subset of the dataset. This validation split, together with the evaluation toolkit, is publicly released for further research. A held-out test split will be reserved for a future shared task focused on document-level translation with formatting preservation.

cs.CL↗

arXiv2026

MEEDAV: A Synchronous Web Viewer for EEG, Eye-Tracking and Speech Data

MEEDAV is an open-source web-based application for the synchronised visualisation of electroencephalography (EEG), eye-tracking, and audio data collected in psycholinguistic research. While originally developed for the Eyetracked Multi-Modal Translation (EMMT) corpus, which uses four-channel EEG data from the Muse 2 headband, MEEDAV also supports higher-density EEG setups thanks to its channel-agnostic processing pipeline. The system performs time alignment across all modalities and provides optional ICA-based EEG denoising. It features interactive Plotly visualisations, including unified EEG-audio-gaze timelines, gaze-intensity plots, event markers, and spatial heatmaps of fixation/saccade patterns. Researchers can filter by participant and stimulus, inspect raw versus cleaned signals, and compute cross-modal correlations. All processing is handled in real time, with a modular backend that supports local file access or GitHub-based streaming. Although initially tailored to the structure of the EMMT dataset, MEEDAV demonstrates a generalisable approach to multimodal data exploration and offers a lightweight, browser-accessible solution for cognitive neuroscience and translation studies.

cs.HC↗

arXiv2026

Training data generation for context-dependent rubric-based short answer grading

Every four years, the PISA test is administered by the OECD to test the knowledge of teenage students worldwide and allow for comparisons of educational systems. However, having to avoid language differences and annotator bias makes the grading of student answers challenging. For these reasons, it would be interesting to consider methods of automatic student answer grading. To train some of these methods, which require machine learning, or to compute parameters or select hyperparameters for those that do not, a large amount of domain-specific data is needed. In this work, we explore a small number of methods for creating a large-scale training dataset using only a relatively small confidential dataset as a reference, leveraging a set of very simple derived text formats to preserve confidentiality. Using the proposed methods, we successfully created three surrogate datasets that are, at the very least, superficially more similar to the reference dataset than a straightforward result of prompt-based generation. Early experiments suggest one of these approaches might also lead to improved training of automatic answer grading models.

cs.CL↗

arXiv2026

Better Late Than Never: Meta-Evaluation of Latency Metrics for Simultaneous Speech-to-Text Translation

Simultaneous speech-to-text translation systems must balance translation quality with latency. Although quality evaluation is well established, latency measurement remains a challenge. Existing metrics produce inconsistent results, especially in short-form settings with artificial presegmentation. We present the first comprehensive meta-evaluation of latency metrics across language pairs and systems. We uncover a structural bias in current metrics related to segmentation. We introduce YAAL (Yet Another Average Lagging) for a more accurate short-form evaluation and LongYAAL for unsegmented audio. We propose SoftSegmenter, a resegmentation tool based on soft word-level alignment. We show that YAAL and LongYAAL, together with SoftSegmenter, outperform popular latency metrics, enabling more reliable assessments of short- and long-form simultaneous speech translation systems. We implement all artifacts within the OmniSTEval toolkit: https://github.com/pe-trik/OmniSTEval.

cs.CL↗

arXiv2025

Corpus of Cross-lingual Dialogues with Minutes and Detection of Misunderstandings

Speech processing and translation technology have the potential to facilitate meetings of individuals who do not share any common language. To evaluate automatic systems for such a task, a versatile and realistic evaluation corpus is needed. Therefore, we create and present a corpus of cross-lingual dialogues between individuals without a common language who were facilitated by automatic simultaneous speech translation. The corpus consists of 5 hours of speech recordings with ASR and gold transcripts in 12 original languages and automatic and corrected translations into English. For the purposes of research into cross-lingual summarization, our corpus also includes written summaries (minutes) of the meetings. Moreover, we propose automatic detection of misunderstandings. For an overview of this task and its complexity, we attempt to quantify misunderstandings in cross-lingual meetings. We annotate misunderstandings manually and also test the ability of current large language models to detect them automatically. The results show that the Gemini model is able to identify text spans with misunderstandings with recall of 77% and precision of 47%.

cs.CL↗

arXiv2025

Finetuning LLMs for EvaCun 2025 token prediction shared task

In this paper, we present our submission for the token prediction task of EvaCun 2025. Our sys-tems are based on LLMs (Command-R, Mistral, and Aya Expanse) fine-tuned on the task data provided by the organizers. As we only pos-sess a very superficial knowledge of the subject field and the languages of the task, we simply used the training data without any task-specific adjustments, preprocessing, or filtering. We compare 3 different approaches (based on 3 different prompts) of obtaining the predictions, and we evaluate them on a held-out part of the data.

cs.CL↗

arXiv2025

End-to-end Automatic Speech Recognition and Speech Translation: Integration of Speech Foundational Models and LLMs

Speech Translation (ST) is a machine translation task that involves converting speech signals from one language to the corresponding text in another language; this task has two different approaches, namely the traditional cascade and the more recent end-to-end. This paper explores a combined end-to-end architecture of pre-trained speech encoders and Large Language Models (LLMs) for performing both Automatic Speech Recognition (ASR) and ST simultaneously. Experiments with the English-to-German language pair show that our best model not only can achieve better translation results than SeamlessM4T, a large foundational end-to-end, multi-modal translation model, but can also match the performance of a cascaded system with Whisper and NLLB, with up to a score gain of 8% in $\text{COMET}^{\text{DA}}_{22}$ metric.

cs.CL↗

arXiv2025

ParCzech4Speech: A New Speech Corpus Derived from Czech Parliamentary Data

We introduce ParCzech4Speech 1.0, a processed version of the ParCzech 4.0 corpus, targeted at speech modeling tasks with the largest variant containing 2,695 hours. We combined the sound recordings of the Czech parliamentary speeches with the official transcripts. The recordings were processed with WhisperX and Wav2Vec 2.0 to extract automated audio-text alignment. Our processing pipeline improves upon the ParCzech 3.0 speech recognition version by extracting more data with higher alignment reliability. The dataset is offered in three flexible variants: (1) sentence-segmented for automatic speech recognition and speech synthesis tasks with clean boundaries, (2) unsegmented preserving original utterance flow across sentences, and (3) a raw-alignment for further custom refinement for other possible tasks. All variants maintain the original metadata and are released under a permissive CC-BY license. The dataset is available in the LINDAT repository, with the sentence-segmented and unsegmented variants additionally available on Hugging Face.

cs.CL↗

arXiv2025

Preliminary Ranking of WMT25 General Machine Translation Systems

We present the preliminary rankings of machine translation (MT) systems submitted to the WMT25 General Machine Translation Shared Task, as determined by automatic evaluation metrics. Because these rankings are derived from automatic evaluation, they may exhibit a bias toward systems that employ re-ranking techniques, such as Quality Estimation or Minimum Bayes Risk decoding. The official WMT25 ranking will be based on human evaluation, which is more reliable and will supersede these results. The official WMT25 ranking will be based on human evaluation, which is more reliable and will supersede these results. The purpose of releasing these findings now is to assist task participants with their system description papers; not to provide final findings.

cs.CL↗

arXiv2025

LLM Compression: How Far Can We Go in Balancing Size and Performance?

Quantization is an essential and popular technique for improving the accessibility of large language models (LLMs) by reducing memory usage and computational costs while maintaining performance. In this study, we apply 4-bit Group Scaling Quantization (GSQ) and Generative Pretrained Transformer Quantization (GPTQ) to LLaMA 1B, Qwen 0.5B, and PHI 1.5B, evaluating their impact across multiple NLP tasks. We benchmark these models on MS MARCO (Information Retrieval), BoolQ (Boolean Question Answering), and GSM8K (Mathematical Reasoning) datasets, assessing both accuracy and efficiency across various tasks. The study measures the trade-offs between model compression and task performance, analyzing key evaluation metrics, namely accuracy, inference latency, and throughput (total output tokens generated per second), providing insights into the suitability of low-bit quantization for real-world deployment. Using the results, users can then make suitable decisions based on the specifications that need to be met. We discuss the pros and cons of GSQ and GPTQ techniques on models of different sizes, which also serve as a benchmark for future experiments.

cs.CL↗

arXiv2025

Overview of the Sensemaking Task at the ELOQUENT 2025 Lab: LLMs as Teachers, Students and Evaluators

ELOQUENT is a set of shared tasks that aims to create easily testable high-level criteria for evaluating generative language models. Sensemaking is one such shared task. In Sensemaking, we try to assess how well generative models ``make sense out of a given text'' in three steps inspired by exams in a classroom setting: (1) Teacher systems should prepare a set of questions, (2) Student systems should answer these questions, and (3) Evaluator systems should score these answers, all adhering rather strictly to a given set of input materials. We report on the 2025 edition of Sensemaking, where we had 7 sources of test materials (fact-checking analyses of statements, textbooks, transcribed recordings of a lecture, and educational videos) spanning English, German, Ukrainian, and Czech languages. This year, 4 teams participated, providing us with 2 Teacher submissions, 2 Student submissions, and 2 Evaluator submissions. We added baselines for Teacher and Student using commercial large language model systems. We devised a fully automatic evaluation procedure, which we compare to a minimalistic manual evaluation. We were able to make some interesting observations. For the first task, the creation of questions, better evaluation strategies will still have to be devised because it is difficult to discern the quality of the various candidate question sets. In the second task, question answering, the LLMs examined overall perform acceptably, but restricting their answers to the given input texts remains problematic. In the third task, evaluation of question answers, our adversarial tests reveal that systems using the LLM-as-a-Judge paradigm erroneously rate both garbled question-answer pairs and answers to mixed-up questions as acceptable.

cs.CL↗

arXiv2025

Intrinsic vs. Extrinsic Evaluation of Czech Sentence Embeddings: Semantic Relevance Doesn't Help with MT Evaluation

In this paper, we compare Czech-specific and multilingual sentence embedding models through intrinsic and extrinsic evaluation paradigms. For intrinsic evaluation, we employ Costra, a complex sentence transformation dataset, and several Semantic Textual Similarity (STS) benchmarks to assess the ability of the embeddings to capture linguistic phenomena such as semantic similarity, temporal aspects, and stylistic variations. In the extrinsic evaluation, we fine-tune each embedding model using COMET-based metrics for machine translation evaluation. Our experiments reveal an interesting disconnect: models that excel in intrinsic semantic similarity tests do not consistently yield superior performance on downstream translation evaluation tasks. Conversely, models with seemingly over-smoothed embedding spaces can, through fine-tuning, achieve excellent results. These findings highlight the complex relationship between semantic property probes and downstream task, emphasizing the need for more research into 'operationalizable semantics' in sentence embeddings, or more in-depth downstream tasks datasets (here translation evaluation)

cs.CL↗

arXiv2025

MockConf: A Student Interpretation Dataset: Analysis, Word- and Span-level Alignment and Baselines

In simultaneous interpreting, an interpreter renders a source speech into another language with a very short lag, much sooner than sentences are finished. In order to understand and later reproduce this dynamic and complex task automatically, we need dedicated datasets and tools for analysis, monitoring, and evaluation, such as parallel speech corpora, and tools for their automatic annotation. Existing parallel corpora of translated texts and associated alignment algorithms hardly fill this gap, as they fail to model long-range interactions between speech segments or specific types of divergences (e.g., shortening, simplification, functional generalization) between the original and interpreted speeches. In this work, we introduce MockConf, a student interpreting dataset that was collected from Mock Conferences run as part of the students' curriculum. This dataset contains 7 hours of recordings in 5 European languages, transcribed and aligned at the level of spans and words. We further implement and release InterAlign, a modern web-based annotation tool for parallel word and span annotations on long inputs, suitable for aligning simultaneous interpreting. We propose metrics for the evaluation and a baseline for automatic alignment. Dataset and tools are released to the community.

cs.CL↗