In 2026, we held the fourth iteration of the SHROOM Shared Task series: SHROOM-Visions (Shared-task on Hallucinations and Related Observable Overgeneration Mistakes in Vision language models), which is hosted at the UncertaiNLP Workshop co-located with EMNLP 2026. Following the success of the 2024 and 2025 tasks, this time we aim to tackle hallucinations through a model-agnostic detection task focused on large vision-language models. Building on the recently introduced SHEEP dataset, designed for long-term evaluation across model generations, the task invites participants to detect and classify fine-grained hallucination spans in image-conditioned text generation (VQA, image captioning, etc.). The evaluation uses a five-class taxonomy of hallucinations spanning four languages: Chinese, English, French, and Italian. The shared task generated strong interest in the NLP community worldwide, with 27 teams contributing 600+ system submissions. The best systems achieve average scores of 0.58 in character-level correlation, 0.46 in label-conditioned correlation, and 0.51 in intersection-over-union (IoU) across four languages, outperforming the baselines by 30–40 points.

This paper introduces OpenSubtitles2024, a massively parallel dataset compiled from translated subtitles. The collection includes an extensive collection of aligned training data based on user-contributed subtitles derived from OpenSubtitles.org and a dedicated held-out dataset for development and evaluation of machine translation and multilingual language models. The collection provides an increased language coverage and doubles the size of the previous edition. Furthermore, a careful procedure was applied to reserve a subset of the most recent subtitles for system development and evaluation. The collection covers 92 languages and language variants, aligned in over 3,000 bitexts containing 40 billion tokens in 7.7 million subtitle files. The test set comprises 2,022 language pairs. In addition, we also provide a multi-parallel test set that refers to a subset of the held-out data with synchronized alignments across 40 languages and 15 subtitles.

This paper investigates a critical design decision in the practice of massively multilingual continual pre-training — the inclusion of parallel data. Specifically, we study the impact of bilingual translation data for massively multilingual language adaptation of the Llama 3 family of models to 500 languages. To this end, we construct a bilingual translation corpus named MaLA, containing data from more than 2,500 language pairs. Subsequently, we develop the EMMA-500 Llama 3 suite of four massively multilingual models — continually pre-trained from the Llama 3 family of base models extensively on diverse data mixes up to 671B tokens — and explore the effect of continual pre-training with or without bilingual translation data. Comprehensive evaluation across 7 tasks and 12 benchmarks demonstrates that bilingual data tends to enhance language transfer and performance, particularly for low-resource languages. We open-source the MaLA corpus, EMMA-500 Llama 3 suite artifacts, code, and model generations.

In an age of rapid model turnover, how do we make hallucination evaluation more perennial? We explore whether human-written hallucination samples could take the place of model-generated hallucinations, in order to make benchmarking detection independent of particular models. To this end, we construct SHEEP, a dataset of 1,600 human-written samples, spanning four languages (Chinese, English, French, Italian), and 18,400 samples from five vision-and-language models, all annotated for hallucinations using a fine-grained span-level labeling scheme. We find that human-written samples result in higher agreement and allow greater control of dataset contents, while remaining distributionally similar to samples derived from vision-and-language samples and providing a reasonable portrayal of detection capabilities — suggesting that human data is a viable substitute for model-based hallucination benchmarks.

Large language models (LLMs) are advancing at an unprecedented pace globally, with regions increasingly adopting these models for applications in their primary languages. Evaluating these models in diverse linguistic environments, especially in low-resource languages, has become a major challenge for academia and industry. Existing evaluation frameworks suffer from inconsistency across different benchmarks, being disproportionately focused on English and a handful of high-resource languages, thereby overlooking the realistic performance of LLMs in multilingual and lower-resource scenarios. To address this critical challenge of fragmented and inconsistent multilingual evaluation, we introduce GlotEval, a unified and lightweight framework that systematically integrates 27 benchmarks under a standardized ISO 639-3 language identifier system, allowing for seamless incorporation of new benchmarks. Supporting nine key tasks (machine translation, text classification, summarization, open-ended generation, reading comprehension, sequence labeling, intrinsic evaluation, instruction following and reasoning), spanning over dozens to hundreds of languages, GlotEval uniquely enables language-specific, cross-benchmark analysis and non-English-centric evaluations at a scale previously less practical for many researchers. This enables a precise diagnosis of model strengths and weaknesses in diverse linguistic contexts. A multilingual translation case study demonstrates GlotEval’s applicability for multilingual and language-specific evaluations.