Prior studies have shown that distinguishing text generated by Large Language Models (LLMs) from human-written one is highly challenging for humans, and often no better than random guessing. To verify the generalizability of this finding across languages and domains, we perform an extensive case study to identify the upper bound of human detection accuracy. Across 16 datasets covering 9 languages and 9 domains, 19 annotators achieved an average detection accuracy of 87.6%, thus challenging previous conclusions. We find that major gaps between human and machine text lie in concreteness, cultural nuances, and diversity. Prompting by explicitly explaining the distinctions in the prompts can partially bridge the gaps in over 50% of the cases. However, we also find that humans do not always prefer human-written text, particularly when they cannot clearly identify its source. We release our dataset, the human labels, and the annotator metadata at https://github.com/xnlp-lab/HumanEval-MGT.

We present the shared task on narrative similarity and narrative representation learning - NSNRL (pronounced "nass-na-rel"). The task operationalizes narrative similarity as a binary classification problem: determining which of two stories is more similar to an anchor story. We introduce a novel definition of narrative similarity, compatible with both narrative theory and intuitive judgment. Based on the similarity judgments collected under this concept, we also evaluate narrative embedding representations. We collected at least two annotations each for more than 1,000 story summary triples, with each annotation being backed by at least two annotators in agreement. This paper describes the sampling and annotation process for the dataset; further, we give an overview of the submitted systems and the techniques they employ. We received a total of 71 final submissions from 46 teams across our two tracks. In our triple-based classification setup, LLM ensembles make up many of the top-scoring systems, while in the embedding setup, systems with pre- and post-processing on pretrained embedding models perform about on par with custom fine-tuned solutions. Our analysis identifies potential headroom for improvement of automated systems in both tracks. The task website includes visualizations of embeddings alongside instance-level classification results for all teams.

Tendem is a hybrid system where AI handles structured, repeatable work and Human Experts step in when the models fail or to verify results. Each result undergoes a comprehensive quality review before delivery to the Client. To assess Tendem's performance, we conducted a series of in-house evaluations on 94 real-world tasks, comparing it with AI-only agents and human-only workflows carried out by Upwork freelancers. The results show that Tendem consistently delivers higher-quality outputs with faster turnaround times. At the same time, its operational costs remain comparable to human-only execution. On third-party agentic benchmarks, Tendem's AI Agent (operating autonomously, without human involvement) performs near state-of-the-art on web browsing and tool-use tasks while demonstrating strong results in frontier domain knowledge and reasoning.

We introduce JEEM, a benchmark designed to evaluate Vision-Language Models (VLMs) on visual understanding across four Arabic-speaking countries: Jordan, The Emirates, Egypt, and Morocco. JEEM includes the tasks of image captioning and visual question answering, and features culturally rich and regionally diverse content. This dataset aims to assess the ability of VLMs to generalize across dialects and accurately interpret cultural elements in visual contexts. In an evaluation of five prominent open-source Arabic VLMs and GPT-4V, we find that the Arabic VLMs consistently underperform, struggling with both visual understanding and dialect-specific generation. While GPT-4V ranks best in this comparison, the model's linguistic competence varies across dialects, and its visual understanding capabilities lag behind. This underscores the need for more inclusive models and the value of culturally-diverse evaluation paradigms.

The rapid proliferation of large language models (LLMs) has increased the volume of machine-generated texts (MGTs) and blurred text authorship in various domains. However, most existing MGT benchmarks include single-author texts (human-written and machine-generated). This conventional design fails to capture more practical multi-author scenarios, where the user refines the LLM response for natural flow, coherence, and factual correctness. Our paper introduces the Benchmark of Expert-edited Machine-generated Outputs (Beemo), which includes 6.5k texts written by humans, generated by ten instruction-finetuned LLMs, and edited by experts for various use cases, ranging from creative writing to summarization. Beemo additionally comprises 13.1k machine-generated and LLM-edited texts, allowing for diverse MGT detection evaluation across various edit types. We document Beemo's creation protocol and present the results of benchmarking 33 configurations of MGT detectors in different experimental setups. We find that expert-based editing evades MGT detection, while LLM-edited texts are unlikely to be recognized as human-written. Beemo and all materials are publicly available.

Current evaluations of mathematical skills in Large Language Models are constrained by benchmarks lacking scope, particularly for multi-modal problems - frequently relying on school-level (Cobbe et al., 2021; Lu et al., 2023; Zhang et al., 2024), niche Olympiadstyle (Fang et al., 2024; Mao et al., 2024), simple quiz format (Yue et al., 2023; Qiao et al., 2024) or relatively small (Lewkowycz et al., 2022) datasets. To address this, we introduce U-MATH, a novel benchmark comprising 1,100 unpublished open-ended university-level problems sourced from current US curricula, with 20% incorporating visual elements. Given the freeform nature of U-MATH problems, we employ LLM judges for solution evaluation and release mu-MATH, a meta-evaluation benchmark composed of 1,084 U-MATH-derived tasks enabling precise assessment of these judges. Benchmarking leading LLMs reveals marked limitations in multi-modal reasoning, with maximum accuracy reaching 93.1% on textual tasks but only 58.5% on visual ones. Furthermore, solution judgment proves challenging, requiring the most advanced models to achieve meaningfully high performance, even still peaking at an imperfect F1-score of 90.1%. We open-source U-MATH, mu-MATH, and all our evaluation code.(1)

This tutorial (https://tum-nlp.github.io/low-resource-tutorial) is designed for NLP practitioners, researchers, and developers working with multilingual and low-resource languages who seek to create more equitable and socially impactful language technologies. Participants will walk away with a practical toolkit for building end-to-end NLP pipelines for underrepresented languages – from data collection and web crawling to parallel sentence mining, machine translation, and downstream applications such as text classification and multimodal reasoning. The tutorial presents strategies for tackling the challenges of data scarcity and cultural variance, offering hands-on methods and modeling frameworks. We will focus on fair, reproducible, and community-informed development approaches, grounded in real-world scenarios. We will showcase a diverse set of use cases covering over 10 languages from different language families and geopolitical contexts, including both digitally resource-rich and severely underrepresented languages.

We present the GenAI Content Detection Task 1 - a shared task on binary machine generated text detection, conducted as a part of the GenAI workshop at COLING 2025. The task consists of two subtasks: Monolingual (English) and Multilingual. The shared task attracted many participants: 36 teams made official submissions to the Monolingual subtask during the test phase and 27 teams - to the Multilingual. We provide a comprehensive overview of the data, a summary of the results - including system rankings and performance scores - detailed descriptions of the participating systems, and an in-depth analysis of submissions.(1)

Recent advances in large language models (LLMs) have introduced the novel paradigm of using LLMs as judges, where an LLM evaluates and scores the outputs of another LLM, which often correlates highly with human preferences. However, the use of LLM-as-a-judge has been primarily studied in English. In this paper, we evaluate this framework in Russian by introducing the Russian Error tyPes Annotation dataset (REPA), a dataset of 1k user queries and 2k LLM-generated responses. Human annotators labeled each response pair expressing their preferences across ten specific error types, as well as selecting an overall preference. We rank six generative LLMs across the error types using three rating systems based on human preferences. We also evaluate responses using eight LLM judges in zero-shot and few-shot settings. We describe the results of analyzing the judges and position and length biases. Our findings reveal a notable gap between LLM judge performance in Russian and English. However, rankings based on human and LLM preferences show partial alignment, suggesting that while current LLM judges struggle with fine-grained evaluation in Russian, there is potential for improvement.

The ELOQUENT lab for evaluation of generative language model quality and usefulness addresses high-level quality criteria through a set of open-ended shared tasks implemented, where possible, to leverage the ability of systems built on language model to assess their own capacity.

The rapid development of AI-driven tools, particularly large language models (LLMs), is reshaping professional writing. Still, key aspects of their adoption such as languages support, ethics, and long-term impact on writers voice and creativity remain underexplored. In this work, we conducted a questionnaire (N = 301) and an interactive survey (N = 36) targeting professional writers regularly using AI. We examined LLM-assisted writing practices across 25+ languages, ethical concerns, and user expectations. The findings of the survey demonstrate important insights, reflecting upon the importance of: LLMs adoption for non-English speakers; the degree of misinformation, domain and style adaptation; usability and key features of LLMs. These insights can guide further development, benefiting both writers and a broader user base.

ELOQUENT is a CLEF lab for evaluating generative language model quality with a focus on such aspects of quality that do not come to the fore with current standard test suites and test collections and to develop and promote new test regimes and methods that fit a multilingual application scenario for generative artificial intelligence. This year is the second year of ELOQUENT. This year's experiment tracks have evolved from the first year: this year we continue challenging the capability of classifiers to distinguish machine-generated from human-authored text; we explore how consistent language models are in responding to value-oriented questions across languages and system settings; we test how accurately language models are able to predict human preferences between variants of generated material; and we investigate how well language models are able provide sensible topical quizzes to fit given target texts.

The current evaluation of mathematical skills in LLMs is limited, as existing benchmarks are either relatively small, primarily focus on elementary and high-school problems, or lack diversity in topics. Additionally, the inclusion of visual elements in tasks remains largely under-explored. To address these gaps, we introduce U-MATH, a novel benchmark of 1,100 unpublished open-ended university-level problems sourced from teaching materials. It is balanced across six core subjects, with 20 Benchmarking leading LLMs reveals marked limitations in multi-modal reasoning, with maximum accuracy reaching 93.1% on textual tasks but only 58.5% on visual ones. Furthermore, solution judgment proves challenging, requiring the most advanced models to achieve meaningfully high performance, even still peaking at an imperfect F1-score of 90.1%.

Warning: this work contains upsetting or disturbing content. Large language models (LLMs) tend to learn the social and cultural biases present in the raw pre-training data. To test if an LLM's behavior is fair, functional datasets are employed, and due to their purpose, these datasets are highly language and culture-specific. In this paper, we address a gap in the scope of multilingual bias evaluation by presenting a bias detection dataset specifically designed for the Russian language, dubbed as RuBia. The RuBia dataset is divided into 4 domains: gender, nationality, socio-economic status, and diverse, each of the domains is further divided into multiple fine-grained subdomains. Every example in the dataset consists of two sentences with the first reinforcing a potentially harmful stereotype or trope and the second contradicting it. These sentence pairs were first written by volunteers and then validated by native-speaking crowdsourcing workers. Overall, there are nearly 2,000 unique sentence pairs spread over 19 subdomains in RuBia. To illustrate the dataset's purpose, we conduct a diagnostic evaluation of state-of-the-art or near-state-of-the-art LLMs and discuss the LLMs' predisposition to social biases.

We present an approach to evaluate the robustness of pre-trained vision and language (V&L) models to noise in input data. Given a source image/text, we perturb it using standard computer vision (CV) / natural language processing (NLP) techniques and feed it to a V&L model. To track performance changes, we explore the problem of visual questions answering (VQA). Overall, we utilize 5 image and 9 text perturbation techniques and probe three Transformer-based V&L models followed by a broad analysis of their behavior and a detailed comparison. We discovered several key findings regarding the performance of the models in relation to the impact of various perturbations. These discrepancies in performance can be attributed to differences in their architectures and learning objectives. Last, but not least, we perform an empirical study to assess whether the attention mechanism of V&L Transformers learns to align modalities. We hypothesize, that attention weights for related objects and words, should be on average higher than for random object/word pairs. However, our study shows that, unlike is believed for machine translation models, V&L models do not learn alignment at all or exhibit less evidence to do so. This may support the intuition that V&L Transformers overfit to either of the modalities.

With recent advances in natural language generation, risks associated with the rapid proliferation and misuse of generative language models for malicious purposes steadily increase. Artificial text detection (ATD) has emerged to develop resources and computational methods to mitigate these risks, such as generating fake news and scientific article reviews. This paper introduces corpus of artificial texts (CoAT), a large-scale corpus of human-written and generated texts for the Russian language. CoAT spans six domains and comprises outputs from 13 text generation models (TGMs), which differ in the number of parameters, architectural choices, pre-training objectives, and downstream applications. We detail the data collection methodology, conduct a linguistic analysis of the corpus, and present a detailed analysis of the ATD experiments with widely used artificial text detectors. The results demonstrate that the detectors perform well on the seen TGMs, but fail to generalise to unseen TGMs and domains. We also find it challenging to identify the author of the given text, and human annotators significantly underperform the detectors. We release CoAT, the codebase, two ATD leaderboards, and other materials used in the paper.

This paper describes AIpom, a system designed to detect a boundary between human-written and machine-generated text (SemEval-2024 Task 8, Subtask C: Human-Machine Mixed Text Detection). We propose a two-stage pipeline combining predictions from an instruction-tuned decoder-only model and encoder-only sequence taggers. AIpom is ranked second on the leaderboard while achieving a Mean Absolute Error of 15.94. Ablation studies confirm the benefits of pipelining encoder and decoder models, particularly in terms of improved performance.

Training and deploying machine learning models relies on a large amount of human-annotated data. As human labeling becomes increasingly expensive and time-consuming, recent research has developed multiple strategies to speed up annotation and reduce costs and human workload: generating synthetic training data, active learning, and hybrid labeling. This tutorial is oriented toward practical applications: we will present the basics of each strategy, highlight their benefits and limitations, and discuss in detail real-life case studies. Additionally, we will walk through best practices for managing human annotators and controlling the quality of the final dataset. The tutorial includes a hands-on workshop, where attendees will be guided in implementing a hybrid annotation setup. This tutorial is designed for NLP practitioners from both research and industry backgrounds who are involved in or interested in optimizing data labeling projects.