Marek Šuppa

Stephen Mayhew, Terra Blevins, Shuheng Liu et al. · cambridge, uw

We introduce Universal NER (UNER), an open, community-driven project to develop gold-standard NER benchmarks in many languages. The overarching goal of UNER is to provide high-quality, cross-lingually consistent annotations to facilitate and standardize multilingual NER research. UNER v1 contains 18 datasets annotated with named entities in a cross-lingual consistent schema across 12 diverse languages. In this paper, we detail the dataset creation and composition of UNER; we also provide initial modeling baselines on both in-language and cross-lingual learning settings. We release the data, code, and fitted models to the public.

Dávid Šuba, Marek Šuppa, Jozef Kubík et al.

Named Entity Recognition (NER) is a fundamental NLP tasks with a wide range of practical applications. The performance of state-of-the-art NER methods depends on high quality manually anotated datasets which still do not exist for some languages. In this work we aim to remedy this situation in Slovak by introducing WikiGoldSK, the first sizable human labelled Slovak NER dataset. We benchmark it by evaluating state-of-the-art multilingual Pretrained Language Models and comparing it to the existing silver-standard Slovak NER dataset. We also conduct few-shot experiments and show that training on a sliver-standard dataset yields better results. To enable future work that can be based on Slovak NER, we release the dataset, code, as well as the trained models publicly under permissible licensing terms at https://github.com/NaiveNeuron/WikiGoldSK.

Marek Šuppa, Viktória Ondrejová, Lucia Ganajová et al.

We present RAGthoven, our system for SemEval-2026 Task 1 (MWAHAHA), Subtask A (multilingual constrained humor generation in English, Spanish, and Chinese). RAGthoven decomposes creative text generation into a multi-stage large language model (LLM) pipeline (Planner, Best-of-N Writer, Reflector for self-critique, LLM-as-a-judge Judge) grounded in computational humor theory (Benign Violation Theory, Script-based Semantic Theory of Humor) and refined across ten experiments. In our final configuration, we augment the Planner with retrieval-augmented generation (RAG) from a curated joke corpus, seeding generation with diverse joke mechanisms. We also evaluate two agentic variants -- ReAct-style sequential tool-calling (Exp09) and autonomous multi-branch orchestration (Exp10) -- that expose the same four stages with a deterministic ConstraintAudit checker. Across four frontier models on a held-out 12-instance English sample, neither agentic variant produced outputs we judged superior to the non-agentic pipeline despite substantially higher tool-call budgets. RAGthoven shares Rank 1 with the Gemini 2.5 Flash baseline in all three languages, with overlapping organizer-reported confidence intervals. In Spanish, it leads the baseline by 42 raw Elo points (1182 vs. 1140), while in English (1045 vs. 1081) and Chinese (1045 vs. 1053) the baseline holds the higher raw rating within the same statistical tie. Together, these results suggest language-dependent diminishing returns from elaborate multi-stage prompt engineering and agentic scaffolding once a strong frontier model is in the loop.

Rishi Desai, Jesse Hu, Joan Cabezas et al.

AI agents are increasingly expected to complete long-horizon workflows that require sustained progress over hours, millions of tokens, and complex environments. Yet current agent benchmarks largely evaluate short-form tasks, such as single pull requests, small tickets, or 5-10 minute exercises, limiting our ability to measure agents' capabilities in planning, long-context understanding, and memory use. We introduce SWE-Marathon, a benchmark of 20 long-horizon tasks spanning software engineering and adjacent technical domains. Each task consists of a unique executable environment, a human-written reference solution, and a multi-layer verification suite. Logged agent attempts average 27.2M total tokens, making SWE-Marathon substantially longer-horizon than existing SWE and command-line agent benchmarks. Current frontier coding agents solve fewer than 30% of tasks. Failures often arise from poor self-verification, self-reported infeasibility, and premature termination. We also observe reward-hacking behavior in 13.8% of rollouts, where agents attempt to exploit the environment or verifier to bypass the intended workflow. SWE-Marathon includes adversarial review of test suites and execution environments, as well as multi-layer checks designed to prevent shortcut solutions. We release SWE-Marathon, evaluation code, and agent trajectories at https://swe-marathon.org/.

Marek Šuppa, Daniel Skala, Daniela Jašš et al.

This study details our approach for the CASE 2024 Shared Task on Climate Activism Stance and Hate Event Detection, focusing on Hate Speech Detection, Hate Speech Target Identification, and Stance Detection as classification challenges. We explored the capability of Large Language Models (LLMs), particularly GPT-4, in zero- or few-shot settings enhanced by retrieval augmentation and re-ranking for Tweet classification. Our goal was to determine if LLMs could match or surpass traditional methods in this context. We conducted an ablation study with LLaMA for comparison, and our results indicate that our models significantly outperformed the baselines, securing second place in the Target Detection task. The code for our submission is available at https://github.com/NaiveNeuron/bryndza-case-2024

Marek Šuppa, Andrej Ridzik, Daniel Hládek et al.

In this work, we introduce skLEP, the first comprehensive benchmark specifically designed for evaluating Slovak natural language understanding (NLU) models. We have compiled skLEP to encompass nine diverse tasks that span token-level, sentence-pair, and document-level challenges, thereby offering a thorough assessment of model capabilities. To create this benchmark, we curated new, original datasets tailored for Slovak and meticulously translated established English NLU resources. Within this paper, we also present the first systematic and extensive evaluation of a wide array of Slovak-specific, multilingual, and English pre-trained language models using the skLEP tasks. Finally, we also release the complete benchmark data, an open-source toolkit facilitating both fine-tuning and evaluation of models, and a public leaderboard at https://github.com/slovak-nlp/sklep in the hopes of fostering reproducibility and drive future research in Slovak NLU.

Marek Šuppa, Ondrej Jariabka, Adrián Matejov et al.

Introductory UNIX courses are typically organized as lectures, accompanied by a set of exercises, whose solutions are submitted to and reviewed by the lecturers. While this arrangement has become standard practice, it often requires the use of an external tool or interface for submission and does not automatically check its correctness. That in turn leads to increased workload and makes it difficult to deal with potential plagiarism. In this work we present TermAdventure (TA), a suite of tools for creating interactive UNIX exercises. These resemble text adventure games, which immerse the user in a text environment and let them interact with it using textual commands. In our case the ''adventure'' takes place inside a UNIX system and the user interaction happens via the standard UNIX command line. The adventure is a set of exercises, which are presented and automatically evaluated by the system, all from within the command line environment. The suite is released under an open source license, has minimal dependencies and can be used either on a UNIX-style server or a desktop computer running any major OS platform through Docker. We also reflect on our experience of using the presented suite as the primary teaching tool for an introductory UNIX course for Data Scientists and discuss the implications of its deployment in similar courses. The suite is released under the terms of an open-source license at \url{https://github.com/NaiveNeuron/TermAdventure}.

Kenneth Enevoldsen, Isaac Chung, Imene Kerboua et al. · cambridge, meta-ai

Text embeddings are typically evaluated on a limited set of tasks, which are constrained by language, domain, and task diversity. To address these limitations and provide a more comprehensive evaluation, we introduce the Massive Multilingual Text Embedding Benchmark (MMTEB) - a large-scale, community-driven expansion of MTEB, covering over 500 quality-controlled evaluation tasks across 250+ languages. MMTEB includes a diverse set of challenging, novel tasks such as instruction following, long-document retrieval, and code retrieval, representing the largest multilingual collection of evaluation tasks for embedding models to date. Using this collection, we develop several highly multilingual benchmarks, which we use to evaluate a representative set of models. We find that while large language models (LLMs) with billions of parameters can achieve state-of-the-art performance on certain language subsets and task categories, the best-performing publicly available model is multilingual-e5-large-instruct with only 560 million parameters. To facilitate accessibility and reduce computational cost, we introduce a novel downsampling method based on inter-task correlation, ensuring a diverse selection while preserving relative model rankings. Furthermore, we optimize tasks such as retrieval by sampling hard negatives, creating smaller but effective splits. These optimizations allow us to introduce benchmarks that drastically reduce computational demands. For instance, our newly introduced zero-shot English benchmark maintains a ranking order similar to the full-scale version but at a fraction of the computational cost.

Jozef Kubík, Marek Šuppa, Martin Takáč

Limited data for low-resource languages typically yield weaker language models (LMs). Since pre-training is compute-intensive, it is more pragmatic to target improvements during fine-tuning. In this work, we examine the use of Active Learning (AL) methods augmented by structured data selection strategies which we term 'Active Learning schedulers', to boost the fine-tuning process with a limited amount of training data. We connect the AL to data clustering and propose an integrated fine-tuning pipeline that systematically combines AL, clustering, and dynamic data selection schedulers to enhance model's performance. Experiments in the Slovak, Maltese, Icelandic and Turkish languages show that the use of clustering during the fine-tuning phase together with AL scheduling can simultaneously produce annotation savings up to 30% and performance improvements up to four F1 score points, while also providing better fine-tuning stability.

Israfel Salazar, Manuel Fernández Burda, Shayekh Bin Islam et al. · mila

The evaluation of vision-language models (VLMs) has mainly relied on English-language benchmarks, leaving significant gaps in both multilingual and multicultural coverage. While multilingual benchmarks have expanded, both in size and languages, many rely on translations of English datasets, failing to capture cultural nuances. In this work, we propose Kaleidoscope, as the most comprehensive exam benchmark to date for the multilingual evaluation of vision-language models. Kaleidoscope is a large-scale, in-language multimodal benchmark designed to evaluate VLMs across diverse languages and visual inputs. Kaleidoscope covers 18 languages and 14 different subjects, amounting to a total of 20,911 multiple-choice questions. Built through an open science collaboration with a diverse group of researchers worldwide, Kaleidoscope ensures linguistic and cultural authenticity. We evaluate top-performing multilingual vision-language models and find that they perform poorly on low-resource languages and in complex multimodal scenarios. Our results highlight the need for progress on culturally inclusive multimodal evaluation frameworks.

Daniel Kyselica, Marek Šuppa, Jiří Šilha et al.

Space debris presents a critical challenge for the sustainability of future space missions, emphasizing the need for robust and standardized identification methods. However, a comprehensive benchmark for rocket body classification remains absent. This paper addresses this gap by introducing the RoBo6 dataset for rocket body classification based on light curves. The dataset, derived from the Mini Mega Tortora database, includes light curves for six rocket body classes: CZ-3B, Atlas 5 Centaur, Falcon 9, H-2A, Ariane 5, and Delta 4. With 5,676 training and 1,404 test samples, it addresses data inconsistencies using resampling, normalization, and filtering techniques. Several machine learning models were evaluated, including CNN and transformer-based approaches, with Astroconformer reporting the best performance. The dataset establishes a common benchmark for future comparisons and advancements in rocket body classification tasks.

Terry Yue Zhuo, Xiaolong Jin, Hange Liu et al.

Crowdsourced model evaluation platforms, such as Chatbot Arena, enable real-time evaluation from human perspectives to assess the quality of model responses. In the coding domain, manually examining the quality of LLM-generated content is extremely challenging, as it requires understanding long chunks of raw code and deliberately simulating code execution. To this end, we introduce BigCodeArena, an open human evaluation platform for code generation backed by a comprehensive and on-the-fly execution environment. Built on top of Chatbot Arena, BigCodeArena enables the execution of LLM-generated code and allows humans to interact with the execution process and outcomes. We collected over 14,000 raw code-centric conversation sessions across 10 widely used LLMs, spanning 10 languages and 8 types of execution environments. Among these conversations, we identified more than 4,700 multi-turn samples with pairwise human preferences. Further analysis uncovers underexplored preferences of LLMs in fine-grained domains characterized by tasks, languages, and frameworks. To systematically examine code understanding and generation capabilities of frontier LLMs, we curated two benchmarks based on the collected data, namely BigCodeReward and AutoCodeArena. For BigCodeReward, we post-processed the 4,700 conversations and evaluated the consistency between reward models and human preferences. The evaluation shows that most LLMs have superior performance in judging coding preferences when the execution results are available. Inspired by these findings, we propose AutoCodeArena, an automatic Elo rating benchmark designed to assess the coding quality of LLMs without human involvement. We find that proprietary LLMs like GPT-5, Claude-Sonnet-4, and Claude-Opus-4 still lead in code generation performance among recent emerging models.

Erik Božík, Marek Šuppa

Automatic Speech Recognition (ASR) for low-resource languages like Slovak is hindered by the scarcity of training data. To address this, we introduce SloPalSpeech, a new, large-scale Slovak ASR dataset containing 2,806 hours of speech from parliamentary proceedings. We developed a robust processing pipeline to align and segment long-form recordings into clean, 30-second audio-transcript pairs suitable for model training. We use this dataset to fine-tune several OpenAI Whisper models (small, medium, large-v3, and large-v3-turbo), achieving significant Word Error Rate (WER) reductions on standard Slovak benchmarks like Common Voice and FLEURS. For instance, the fine-tuned Whisper-small model's WER dropped by up to 70\%, approaching the baseline performance of the much larger Whisper-large-v3 model. To foster future research in low-resource speech recognition, we publicly release the complete SloPalSpeech dataset, the fully segmented transcripts (60 million words), and all our fine-tuned models.

Daniel Kyselica, Tomáš Hrobár, Jiří Šilha et al.

The characterization and analysis of light curves are vital for understanding the physical and rotational properties of artificial space objects such as satellites, rocket stages, and space debris. This paper introduces the Light Curve Dataset Creator (LCDC), a Python-based toolkit designed to facilitate the preprocessing, analysis, and machine learning applications of light curve data. LCDC enables seamless integration with publicly available datasets, such as the newly introduced Mini Mega Tortora (MMT) database. Moreover, it offers data filtering, transformation, as well as feature extraction tooling. To demonstrate the toolkit's capabilities, we created the first standardized dataset for rocket body classification, RoBo6, which was used to train and evaluate several benchmark machine learning models, addressing the lack of reproducibility and comparability in recent studies. Furthermore, the toolkit enables advanced scientific analyses, such as surface characterization of the Atlas 2AS Centaur and the rotational dynamics of the Delta 4 rocket body, by streamlining data preprocessing, feature extraction, and visualization. These use cases highlight LCDC's potential to advance space debris characterization and promote sustainable space exploration. Additionally, they highlight the toolkit's ability to enable AI-focused research within the space debris community.

Matej Králik, Marek Šuppa

Hand Gesture Recognition (HGR) based on inertial data has grown considerably in recent years, with the state-of-the-art approaches utilizing a single handheld sensor and a vocabulary comprised of simple gestures. In this work we explore the benefits of using multiple inertial sensors. Using WaveGlove, a custom hardware prototype in the form of a glove with five inertial sensors, we acquire two datasets consisting of over $11000$ samples. To make them comparable with prior work, they are normalized along with $9$ other publicly available datasets, and subsequently used to evaluate a range of Machine Learning approaches for gesture recognition, including a newly proposed Transformer-based architecture. Our results show that even complex gestures involving different fingers can be recognized with high accuracy. An ablation study performed on the acquired datasets demonstrates the importance of multiple sensors, with an increase in performance when using up to three sensors and no significant improvements beyond that.

Katarína Benešová, Andrej Švec, Marek Šuppa

In this study we demonstrate the viability of deploying BERT-style models to serverless environments in a production setting. Since the freely available pre-trained models are too large to be deployed in this way, we utilize knowledge distillation and fine-tune the models on proprietary datasets for two real-world tasks: sentiment analysis and semantic textual similarity. As a result, we obtain models that are tuned for a specific domain and deployable in serverless environments. The subsequent performance analysis shows that this solution results in latency levels acceptable for production use and that it is also a cost-effective approach for small-to-medium size deployments of BERT models, all without any infrastructure overhead.