Multilingual Language Model Evaluation

Document-level text simplification involves transformations that go beyond sentence-internal edits, addressing discourse coherence, anaphora resolution, and cross-paragraph consistency. Despite advances in sentence-level simplification for high-resource languages, document-level simplification in morphologically rich, low-resource languages such as Estonian remains largely unexplored. This study presents a comprehensive evaluation of five state-of-the-art multilingual large language models (LLMs) for document-level simplification in Estonian. Three prompting strategies are examined: single-pass generation, pipeline-based modular agents, and guideline-augmented pipelines. The evaluation framework integrates automatic metrics assessing readability, semantic preservation, and discourse coherence, alongside a structured manual annotation protocol. The findings indicate that Gemini-2.0 and LLaMA-3.3 produce outputs with near-native fluency and strong meaning preservation, whereas other models display notable grammatical and semantic limitations. This work contributes novel document-level coherence metrics, evidence-based prompting strategies, and publicly available resources for reproducibility.

Skeleton-based reasoning prompting is a promising training-free approach for structuring LLM reasoning, but prior work largely assumes an English-centric setting. We propose the Language-Aware Skeleton Exploration Framework (LASEF) to study skeleton-language choice in multilingual mathematical reasoning. Across math benchmarks, model scales, and languages, we show that English skeletons yield a small positive tendency on average, most visible for smaller models and low-resource languages. However, few language-level gains remain significant after correction, and English is not universally optimal. Combining greedy decoding, multi-rollout evaluation, translation ablation, and cross-benchmark validation, we further find three patterns of skeleton-language effects: directionally consistent, evaluation- and benchmark-dependent, and asymmetric negative. These effects cannot be fully explained by generation quality alone. Overall, skeleton language is a context-dependent design variable that requires multi-level exploration. All resources are released at https://github.com/lhsstn/LASEF.

Multilingual LLM benchmarks omit Slovak, a morphologically rich West Slavic language of five million speakers, or cover it only by machine translation. We present sk-bench, a native-first Slovak benchmark with 30 datasets (33 scored task variants) across ten skill categories. Eleven resources are introduced or first packaged for generative-LLM evaluation, including IFEval-SK with Slovak-adapted instruction checkers and native Chiby/SKJ1 resources for Slovak grammar and morphology. We evaluate 55 open- and closed-weights models under one harness. The best open model trails proprietary APIs by 12.6 points. Model rankings are similar for native and translated closed-form data ($ρ\geq0.98$), though translation separates the strongest models less well. By contrast, human-authored and LLM-generated QA questions rank models differently ($ρ=0.72$). For Qwen3-14B, continued Slovak pretraining lowers the overall score by 13.9 points. A small instruction set restores three quarters of that loss. Test-time reasoning improves scores by 8.5 to 12.5 points for models of 9B and above. Together, these findings suggest four design lessons for other under-resourced languages: use native data where translation fails, plan instruction repair after language adaptation, enable test-time reasoning before scaling up, and avoid overinvesting in target-language prompts. We release the data and code at https://github.com/slovak-nlp/sk-bench

A common approach to measuring bias in Large Language Models is to compare the log-likelihoods of two contrastive stereotype sentences. We argue that such single-pair comparisons are often unreliable: simply rewriting the same stereotype with an alternative attribute can yield logically inconsistent preferences. To address this, we propose a dual minimal pair setup that introduces two axes of comparison for robust stereotype evaluation. First, we present a data-augmentation framework that fills critical gaps in existing stereotype datasets by generating paraphrases and alternate attributes. We apply our framework on a set of English, Russian, Spanish and Chinese stereotypes. Second, we introduce two evaluation metrics tailored to the dual minimal pair setup. One of these metrics provides a new perspective on bias by modeling the mutual information (MI) between social groups and stereotyped attributes. This MI-based metric is better suited for aggregation and enables more robust comparisons of stereotype strength across different languages and models. Our code is available at https://github.com/stepanat/missing-minimal-pair/.

LLMs are increasingly deployed as autonomous agents in social environments, making it critical to study their ability to faithfully simulate human interactions. Central to this is grounding agents in realistic user personas, yet existing datasets rely on fictional personas and are limited to a handful of languages, lacking the empirical grounding necessary to evaluate behavioral fidelity across diverse populations. We introduce Wiki-Talkie, a multilingual dataset of real-world conversations from Wikipedia Talk pages across five languages spanning two language families: Germanic (German, English) and Romance (Spanish, French, Italian), paired with personas derived from real user communities and encompassing sociodemographic attributes, self-descriptions, and behaviorally grounded interaction traits. Using Wiki-Talkie, we evaluate agent interactional behavior on a next-turn generation task across various persona conditioning strategies. Our evaluation assesses whether agents collectively reproduce the distributional behavioral patterns observed in human discussions. Results show that user's comment history exemplifying interaction behavior consistently outperforms explicit persona information. In addition, models systematically underproduce negative or extreme sentiments, while over producing references and suggestions, revealing biases toward agreeableness and positivity. Crucially, these patterns hold robustly across languages, with small cross-lingual differences.

Current evaluation of multilingual Large Language Models (LLMs) rests on an implicit Translation-Isomorphism Assumption (TIA): that semantic structures across languages are congruent and mutually mappable without loss of information. We argue that this assumption is not merely violated in practice, but ill-posed in principle for a typologically identifiable class of concepts, including pragmatic markers, honorifics, and diachronically stratified terms. We formalize this failure using a usage-cloud framework, representing concepts as point sets of contextualized embeddings. We define $α$-unalignability as the impossibility of any mapping that simultaneously preserves lexical faithfulness (centroid correspondence) and structural faithfulness (local neighborhood topology). We provide three layers of evidence. Behaviorally, we show that FLORES-200 translation failures are predicted by language family and resource class but not by script, and that LOBSTER reasoning scores vary by family. Mechanistically, we report a Representation-Intervention Gap (RIG) in a nine-model case study on Yami: the models' activations encode a regularity along which Yami groups with other low-resource and Austronesian languages, yet interventions on language-specific neurons show no demonstrated advantage over random masks: the regularity is visible but not usable by this intervention. Finally, we operationalize these findings into a multidimensional diagnostic profile: Cycle-Consistency, Pragmatic-Load Disagreement, Manifold-Curvature Mismatch, and RIG. We argue that collapsing cultural competence into a single scalar incentivizes "probabilistic flattening," and that recognizing the unalignable class is a precondition for AI that respects, rather than erases, cultural divergence. This suggests that multilingual alignment is not a single well-defined objective, but a set of mutually incompatible projections.

Language models frequently generate outputs in unintended languages or scripts, a phenomenon known as off-target generation. While existing research has focused on language selection, the dimension of script knowledge remains understudied: before any linguistic understanding can occur, users must recognize the graphic symbols in a model's response. We investigate whether Small and Large Language Models (SLMs and LLMs) possess script knowledge by testing them on multi-scriptic languages. Through two complementary experiments, we evaluate whether models (1) adapt their output script to match the input, and (2) follow explicit instructions to generate text in a specified script. The models we tested demonstrate substantial script knowledge: they all achieve a near-perfect Latin script fidelity (more than 98%) and follow script instructions with high frequency. Nevertheless, we notice differences between LLMs and SLMs, with higher scores for LLMs including for non-standard script combinations.

We understand little about how capabilities acquired in one language carry over to another, or what governs this transfer: evaluations rely on incomparable, saturation-prone datasets and rarely examine its determinants jointly. Identifying what predicts transfer would let us avoid exhaustive evaluation across all language pairs and let developers target the factors that limit performance in low-resource languages. To evaluate cross-lingual capability transfer, we introduce Multilingual GSM-Symbolic, an extensible multilingual mathematical dataset covering 30,000 item-matched question-answer pairs and spanning 15 languages. It utilises symbolic templates to prevent overfitting and ensure generalisation by allowing generation of millions of high-quality variations from a single sample. Using Multilingual GSM-Symbolic, we quantify the largest determinants of capability as model size ($β= 1.77$), language resource level ($β= 0.77$), reasoning ($β= 0.67$) and typological distance ($β= -0.25$). This joint estimation allows these determinants to be expressed in terms of one another: a 32B model evaluated in Marathi performs like a 10B model in English. Our findings have important implications for model developers, showing that model size and reasoning narrow the performance gap between low- and high-resource languages ($β= -0.27$ and $β= -0.20$, respectively), while similar levers have little or no effect on typologically distant languages. Overall, our analysis framework explains 92% of between-language variation, but only 23% of the model-by-language variation, and predicts a model's performance on an unseen language within 6.0pp (r=.96). Incorporating measurements from just 10 templates in the target language reduces this to 4.19pp, enabling reasonable estimates of performance with little or no downstream dataset.

Semantic IDs (SIDs) compress item embeddings into discrete code sequences used in generative retrieval. We ask whether a multilingual encoder is sufficient for different-language renderings of the same product to receive language-consistent SIDs. Using Amazon ESCI listings rendered in English, Spanish, and Japanese, we test whether translations remain close to their English source, whether residual quantization is unusually sensitive to translation-induced movement, and whether multilingual or language-balanced quantizer fitting improves SID agreement. Multilingual E5 places translations measurably apart: under an English-heavy fit, a Japanese translation preserves the first SID code of its English counterpart in only 7.7% of cases, compared with 89.0% for an English rewording. Distance-matched product-directed controls produce nearly the same full-SID mismatch as translation, providing no evidence that the quantizer selectively amplifies language directions. Balancing the fitting mixture makes codebook use more uniform but further reduces cross-lingual prefix agreement: Spanish first-code consistency falls from 28.3% to 6.6%, while an English-only fit preserves it for 67.6% of Spanish translations. These results show that multilingual exposure and balanced codebook use alone do not guarantee language-consistent SIDs.

Large Language Models increasingly serve as interfaces for knowledge-intensive information seeking tasks across languages by synthesizing multilingual evidence. Prior work has shown that they often exhibit query-language preference -- the tendency to favor sources written in the language of the query -- but has largely examined this behavior in settings where equivalent knowledge is available across languages. However, this bias becomes consequential when sources in different languages provide incomplete or inconsistent accounts of the same fact, since the information users receive then depends on the sources a model selects to use. To characterize query-language preference under such cross-lingual knowledge disparities, we introduce Waldo, a multilingual Question-Answering (QA) benchmark constructed from Wikipedia. Waldo contains 12K QA pairs targeting knowledge gaps, where a fact is available in one language but absent in another, and knowledge conflicts, where language editions provide conflicting versions of the same fact. Evaluating eight models across five languages, we find that when one language edition merely lacks the relevant fact, models generally use evidence from the other language regardless of the query language. Under conflicting accounts, however, model responses strongly align with the document in the query language, causing semantically equivalent queries to elicit different accounts depending on the user's language. Finally, we explore two different approaches that could mitigate this preference under knowledge conflicts: a mechanistic intervention that ablates attention heads associated with query-language preference, and LoRA-based training, which reduces the preference gap by up to 61.5%.

Claims about the grammatical competence of multilingual language models vary sharply with how competence is measured, yet the interaction between evaluation paradigm, post-training, and language resource availability has not been systematically examined. We evaluate base and post-trained models from six families on MultiBLiMP, a syntactic minimal-pair benchmark covering 101 languages, using four evaluation methods. We report three principal findings. First, post-training degrades grammatical competence, but the magnitude of this effect is reduced unevenly by model scale, while low-resource languages bear the highest cost. Second, post-trained models retain grammatical knowledge they cannot articulate through explicit prompting, yet this is measurable only in high-resource languages, because near-chance baselines in low-resource settings leave little knowledge to hide. Third, native-language prompting recovers otherwise hidden competence on low-resource languages, demonstrating that only high-resource languages can be probed directly from unprompted probabilities. We conclude that multilingual grammatical evaluation must adopt language-informed, multi-paradigm protocols to avoid systematically underestimating low-resource abilities.

Jev is a commercial System One model from TypeSafe AI that does not generate text: given a state and typed questions, it returns a choice from fixed options, a position on a rubric, or the probability that a statement is true, with probabilities the vendor describes as calibrated. Such models target small decisions in information access pipelines, such as routing queries, checking grounding, moderating content, or rating against a rubric. We evaluate Jev (jev-1.13.0) zero-shot on 37 datasets spanning classification, routing, natural language inference, reading comprehension, commonsense reasoning, moderation, legal clause analysis and rubric scoring, with one frozen template per dataset and full evaluation splits: 346,009 requests for under USD 10. For reference, we score Qwen3.8-27B and Gemma-4-E4B on identical requests via their exact next-token probabilities over the options. Jev reaches 95-99% accuracy on IMDB, SST-2, HellaSwag and ARC and 86.7% on Belebele across 122 languages. It beats Qwen on 27 of 37 datasets, with none of Qwen's nine leads outside the bootstrap intervals, and Gemma on all 37. All three models degrade on low-resource languages, fine-grained or noisy labels, and rubric-based quality judgments. Jev's choice probabilities are well calibrated and support selective prediction. Binary probabilities rank well but are poorly placed relative to a fixed 0.5 threshold; thresholds tuned on training data raise micro-F1 on UNFAIR-ToS from 0.50 to 0.75. Jev answers MMLU's calculation-heavy questions more accurately than other MMLU questions (94% vs. 91%), whereas both open models, and all three on C-Eval, find them harder. Rotating the options leaves Jev's accuracy unchanged and withholding the question drops it to near chance, ruling out shallow memorization but not memorized question-answer pairs. We release the code, harness and all raw responses.

Cross-lingual zero-shot transfer and multilingual fine-tuning are promising approaches for NLP tasks such as Named Entity Recognition (NER) in low-resource languages, but in the absence of target language benchmarks, it is unclear which auxiliary language selection strategy leads to the best transfer. We introduce RunyaNER, the first publicly available NER benchmark for the East African language Runyankore, and use it to investigate the choice of which languages to use for transfer. Created with a semi-automated pipeline and fully manually verified, RunyaNER contains over 237k annotated words across 30k sentences. We benchmark pretrained models on RunyaNER, establishing that our dataset is of sufficient quality and size to produce effective Runyankore NER models. We then use RunyaNER to investigate auxiliary language selection in cross-lingual zero-shot and multilingual fine-tuning settings. Our experiments show that while transfer performance is highly sensitive to auxiliary language selection, embedding-based measures computed from labelled training spans correlate more strongly with downstream transfer performance than traditional linguistic features based on metadata or typology. By releasing RunyaNER and providing a systematic analysis of auxiliary language selection strategies, this work contributes both a new benchmark resource and practical insights for multilingual transfer in low-resource settings.

Open-source reasoning models provide unequal access to reasoning capability across languages. When a model can solve a problem but cannot deliver a complete solution in the user's language, language becomes an access barrier rather than merely a source of performance variation. We audit Qwen3 checkpoints on competition-mathematics tasks in eleven languages. Across the ten non-English languages, only 15.4-17.9% of problems receive a correct, terminating solution with visible reasoning in the requested language in any of 16 samples, compared with 92.9% in English. To identify the source of this disparity, we evaluate thirteen endpoints from one model family, spanning released checkpoints, multilingual supervised fine-tuning (SFT) at two scales, controlled SFT ablations, and three reinforcement-learning (RL) reward formulations. We jointly track correctness, language adherence, termination, and delivery efficiency. The dominant bottleneck shifts across post-training stages. Released models often reason in English. Multilingual SFT restores target-language reasoning, but accuracy declines across multilingual, English-only, and single-language SFT runs, showing that this cost is not specific to multilingual mixing; non-English reasoning traces additionally become prone to non-terminating loops. RL restores termination in both arms at no cost in accuracy, but only the arm whose reward includes a language term delivers: rewarding correctness alone returns the model to English. Together, these stages establish a constructive post-training path from English-pivoted capability to multilingual reasoning that is reliably delivered.

Decision-only language models return a probability for every answer option instead of generating text, which makes them attractive as survey respondents and as judges. We audit the cultural values of one such model, TypeSafe's JEV, with the Values Survey Module 2013. We asked it the 24 items as 12 matched Saudi and 12 matched American personas and without a persona, in English and Arabic, under eight ways of formulating the request (288,000 answers). JEV's answers were highly repeatable (ICC 0.997), and without a persona they resembled those of its own American personas. When the persona was Saudi rather than American, the answers moved in the direction of the human Saudi-US difference, reproducing 87% of its size in English but 62% in Arabic, with long-term orientation reversed. A language cross shows that the smaller difference in Arabic comes from the language of the items, not from the language of the persona description. Age shifted the profiles about as much as nationality, gender shifted them more for Saudi than for American personas, and JEV was less confident in Arabic and for Saudi personas. These patterns held in every request design, although the model never generates text.

Large language models are increasingly evaluated in specialized domains such as law, medicine, software engineering, and cybersecurity, yet film remains comparatively underexplored despite requiring long-form narrative integration, multilingual interpretation, and culturally situated audience judgments. We introduce CineSubBench, a benchmark for evaluating long-context film understanding from multilingual movie subtitles. A subtitle track represents a film as thousands of short, temporally ordered utterances from which models must reconstruct characters, relationships, events, causal progression, and themes without explicit scene or event structure. CineSubBench contains 1,012 films with complete subtitle coverage in six languages, yielding 6,072 tracks and 8.13M timestamped subtitle entries. It provides a matched multi-task, multilingual, and multicultural (MultiX) evaluation setting: seven tasks span narrative reconstruction and abstraction, genre prediction, age suitability, country-specific motion-picture ratings across ten national classification systems, and subtitle-grounded language safety. Across nine LLMs, plot premises are recovered more reliably than event-complete synopses; cross-lingual consistency varies substantially across models and languages; national rating systems expose distinct calibration patterns; and strong profanity is far easier to ground than mild obscenity. CineSubBench establishes film as a long-context LLM evaluation domain and provides a unified benchmark for measuring narrative, multilingual, cultural, and evidence-grounding capabilities.

The rapid advancement of Large Language Models (LLMs) imposes a thorough evaluation of their linguistic and analytical capabilities as well as constraints, particularly for a language with limited benchmark coverage such as Greek. To address the limited availability of comprehensive benchmarks in this domain, we introduce Prot-Ex and Pan-Ex, two benchmarks consisting of questions from entrance exams for Greek Model and Experimental schools as well as the Panhellenic exams (the Greek national university entrance examinations). These benchmarks are employed to assess the performance of text-only LLMs-including the Greek-adapted KriKri-8B-Instruct, Llama-3.1-8B, Gemma-4-26B, and Qwen-3-32B-across diverse academic disciplines (Modern Greek, Mathematics, Physics, etc.) and task formats (closed, structured, and open-ended), including textualized visual context (i.e., image descriptions). Our findings indicate the localized KriKri-8B significantly outperforms its base model, successfully rivalling much larger LLMs in linguistically demanding humanities tasks. By leveraging an LLM-as-a-Judge methodology, we expose the inadequacy of traditional lexical metrics for evaluating complex reasoning. Crucially, we uncover a few-shot prompting paradox: while synthetic examples improve accuracy in closed-ended questions, they severely overload the context window of 8B models in structured tasks, causing significant performance degradation. Ultimately, this study suggests targeted linguistic adaptation offsets lower parameter counts in specialized domains, despite the fragility of smaller models to prompt verbosity.

Misinformation on social media remains a critical problem, and more and more people settle it by asking a language model instead of a fact checker. Whether models judge such claims reliably is debated; whether they judge them equally well in every language people ask in has gone almost unasked. We test eight models from five families, 3B to 70B, on 1,500 encyclopedic factual claims that exist in identical form in eight languages. English is judged better than every other language on every model, and the gap is widest on the smallest ones, where Llama-3B on Arabic is no better than guessing. Existing remedies retrain on more multilingual data or fit an unconstrained map between language representations, and neither asks whether the model already holds the answer and simply fails to say it. It largely does: a linear probe recovers the truth from the very activations the model fails to express. We propose RoSh, a per-language shift and rotation of the residual stream, computed in closed form at three layers, with no training and no weight modified. It improves every model and closes 75% of the gap on average, helping most where the model was worst: Arabic on Llama-3B goes from chance to nearly the English level, and a fifth fewer of the claims answered correctly in English are lost in translation. What remains is no longer a read-out failure: afterwards the head recovers as much of what is encoded outside English as it does in English. An unconstrained map fitted on the same pairs falls below the untouched baseline, so the orthogonality constraint is doing the work, and every model clears a scrambled-correspondence control and ten further controls. On the two benchmarks of the closest inference-time method, latent-space intervention, run with its own data and metric code, RoSh's gains are five to thirteen times larger.

We audit a multilingual affective generation benchmark eight instruction-tuned LLMs producing emoji summaries for 17,100 Bangla, English and Hindi sentences, with 6,960 human judgements and find its headline conclusions to be artefacts of the measurement instrument rather than properties of the systems. Treating annotators as a random rather than a fixed factor, no system differs significantly from any other ($F(7,14)=0.59$, $p=0.76$), although the conventional analysis declares 19 of 28 pairwise differences significant. Annotator identity explains far more rating variance than system identity, and the winning system changes whenever any single annotator is removed. The ordering that does emerge tracks output length: mean emoji count explains 78.7% of between-system variance, and a within-item length-matched comparison over 2,599 pairs reverses the leaderboard. We further show that cross-provider anisotropy differences vanish under mean-centring, that per-language token costs change sign with the normalising unit, and that multi-view row-wise splits inflate macro-F1 by $3.1$ points and change the top-ranked system. In place of preference scoring we propose emoji-affect decodability, a reference-based probe whose rankings are stable to $\pm0.003$ macro-F1 across seeds.

Medical Named Entity Recognition (NER) for low-resource languages remains a challenging task due to high linguistic variability and a scarcity of domain-specific annotated corpora. This work presents a comprehensive empirical benchmark evaluating three fine-tuned transformer encoders-BanglaBERT, multilingual BERT (mBERT), and XLM-RoBERTa-against GPT-4o mini under zero-shot and few-shot prompting configurations for Bangla medical NER. In contrast to prior studies that evaluated large language models on limited subsets of only 50 samples, we conduct a large-scale evaluation across the full test set of 3,179 samples, providing statistically robust and reproducible baselines. Our fine-tuned XLM-RoBERTa model achieves an F1- score of 0.5959, establishing a new state-of-the-art and surpassing the previously reported best result of 0.5848. Crucially, we demonstrate that the language-specific BanglaBERT model consistently underperforms its multilingual counterparts with an F1-score of 0.4937, indicating that pretraining domain diversity can outweigh language specificity in highly specialized clinical settings. Furthermore, we present a detailed per-entity-type analysis for this task, revealing that Medicine and Specialist categories are recognized with high reliability, achieving F1- scores above 0.83, while the Symptom category remains the most challenging with an F1-score of 0.4367 despite being the most frequent training class. Finally, fine-tuned transformer models outperform the optimal prompting configuration by a factor of 3.76, confirming that prompt-only pipelines remain inadequate for structured clinical entity extraction in low-resource language environments.

Most Turkish-capable large language models (LLMs) are evaluated using general-purpose benchmarks rather than long, structurally complex domain documents. This paper evaluates five open-weight 7B-8B models for Turkish document question answering under a resource-constrained local deployment setting. The primary benchmark contains 100 systematically validated questions derived from a 109-page industrial R&D report, and the evaluation protocol is replicated using a second 112-page public-sector report and an independently constructed 100-question set. All models are evaluated locally on an NVIDIA RTX 3050 laptop GPU with 6 GB VRAM using controlled prompting, decoding, and 4-bit quantisation. The principal methodological contribution is an evidence-annotated evaluation protocol that separates retrieval failure from downstream model reasoning failure without requiring additional model calls. On the primary benchmark, end-to-end accuracy ranges from 49% to 75%. Seven lexical, dense, and hybrid retrieval configurations are additionally compared using 95% Wilson intervals and exact paired McNemar tests; none significantly outperforms the character TF-IDF baseline on either document. Evidence recall saturates differently across the two reports, showing that retrieval and effective context capacity can be binding constraints for some documents but not others. These results demonstrate that model selection, retrieval behaviour, and hardware limits must be evaluated separately when deploying open-weight LLMs for Turkish domain documents.

Long-context evaluations often test whether a model can recover distant evidence, but recoverability does not guarantee behavioral influence. We test the prediction that a distant discourse anchor can remain explicitly recoverable yet fail to change the locally preferred reading of a familiar multiword expression; such failures should concentrate when the model's no-anchor default conflicts with the anchor, while prior-correct decisions remain largely preserved. We introduce Multiword Expression Effective Context Length (MWE-ECL), a bilingual diagnostic whose matched anchor-retrieval, no-anchor prior, and interpretation prompts measure explicit recoverability, model-observed defaults, and anchor-conditioned decisions, respectively. Across eight English deployment panels on a shared 0-128K grid, retrieval-control accuracy on prior-conflict items is 0.989-1.000, prior-conflict override spans 0.806-1.000 (0.809-1.000 after conditioning on correct retrieval), and preservation of prior-correct decisions remains 0.977-1.000. A same-call control querying retrieval and interpretation in one prompt reproduces the gap for DeepSeek V4 Pro (1.000 retrieval versus 0.900-0.920 interpretation), showing that separate invocations are not its sole explanation; smaller or absent gaps in the other two models bound its generality. For DeepSeek V4 Flash, separate prompt-fit tests retain perfect retrieval with lower interpretation at 512K and 1M, while foil-consistent cues shift the no-anchor prior far more than retrieval; cross-model cue effects are heterogeneous. A separately reported 10-family Chinese subset shows similar descriptive gaps, but imperfect retrieval for some models prevents an integration-only attribution. MWE-ECL therefore evaluates whether explicitly recoverable distant context changes a competing local semantic decision.

This paper summarizes the Interspeech2026 second Multilingual Conversational Speech Language Model (MLC-SLM) Challenge, which aims to advance the development of effective multilingual conversational speech language models. We describe the two challenge tasks: multilingual conversational speech diarization and recognition, and multilingual conversational speech understanding, together with the released real-world conversational speech dataset, evaluation protocols, and baseline systems. The challenge attracted 91 teams worldwide, with 704 valid leaderboard results and 14 technical reports across the two tasks. Based on the participating systems, we summarize representative approaches and distill practical insights into multilingual conversational speech recognition and understanding to support future research in the community.

Current literature evaluates large language models (LLMs) on multilingual kinship understanding using multiple choice benchmarks, treating it as a recognition problem. We instead prompt five open weight LLMs to generate kinship terms in three non Western languages (Hindi, Tamil, and Korean) across two communicative tasks and pair this with a matched option-supported selection baseline. On identical relation language cells, GPT OSS120B selects the correct term in 90.67% of 75 valid cells but produces an accepted term in 36.00% of the corresponding attempts; Llama 3.370B shows the same pattern (77.92% versus 24.24%). Since the four-option condition displays the candidate terms and does not require script production, the difference is interpreted as an evaluation format gap rather than direct proof that lexical knowledge is intact. On explicitly specified L3 prompts, accuracy varies sharply, from GLM-5.1 at 72.29% to Llama-3.370B at 24.24%. The paternal-lineage advantage is language specific; it is large in Hindi but weak or reversed in Korean, while Tamil shared-term pairs provide a control for measurement variation. These results show that culturally specific kinship generation remains difficult even when the relationship is explicitly stated and motivate generation-based evaluation alongside multiple-choice testing.

Cultural evaluation coverage and robustness in language models are difficult to diagnose because pretraining corpora and cultural benchmarks are rarely indexed with comparable metadata. Benchmarks increasingly target culturally situated phenomena at the level of languages, regions, and locale-specific practices, while web-scale corpora are usually organized only by language. A shared culture-language-region layer makes these resources comparable, enabling audits of whether a target cultural phenomenon is represented in pretraining data, evaluated by benchmarks or both. To this end, we introduce FineWeb-CLaR, a large-scale annotated dataset derived from FineWeb and FineWeb-2 that places web documents on a shared culture-language-region axis for corpus auditing and benchmark alignment. FineWeb-CLaR annotates the full 30.9B-document collection from FineWeb and FineWeb-2 with URL-derived region labels and cultural-topic provenance. Our region resolver assigns a non-empty region to 25.61% of documents (7.92B). For cultural-topic analysis, we induce locale-specific topics and project them onto the 14 leaves of the Cultural Taxonomy of Liu et al. (2025), producing Locale Topic Distributions (LTDs) for corpus-side comparison. We also annotate 277 cultural NLP benchmarks with the same taxonomy, language coverage, and region coverage. Together, these resources enable direct comparison between corpus-side pretraining evidence and benchmark-side evaluation coverage.

Large language model (LLM) agents increasingly execute multi-step workflows through tool use and interaction with users and environments. However, current agent evaluations are largely English-centric, limiting our understanding of agent capabilities in multilingual settings. We introduce BabelFlow, a benchmark-general agentic workflow that adapts existing agent benchmarks to new languages by analyzing runtime dependencies, coordinating structure-preserving translation, and combining multi-layer verification with human review to preserve task and evaluation semantics. Using BabelFlow, we construct BabelArena, a task-aligned benchmark comprising 16,146 instances derived from 702 canonical tasks across four benchmark families, 13 domains, and 23 languages. Experiments with five frontier models show that no single model dominates across benchmark families and that cross-language disparities extend well beyond task success. Lower-resource languages exhibit distinct failure patterns, with larger shares of tool-use and control-flow errors rather than answer-quality errors alone, pointing to gaps in reliable task execution across the resource levels of these languages. On the same tasks, agents in low-resource languages also consume substantially more tokens than in English (up to roughly twice the input) without proportional increases in interaction length, and language consistency degrades further on tasks requiring structured output, where switches are directed overwhelmingly toward English. We believe BabelArena provides a foundation for advancing research on reliable and efficient multilingual agents.

Large language models are increasingly used in healthcare communication, yet most evaluations emphasize response quality while assuming that the user's concern has been interpreted correctly. We introduce HerHealthEval, a controlled evaluation framework for multilingual understanding of women's-health communication. For each clinical case, HerHealthEval provides matched versions in English, French, and Modern Standard Arabic using six communicative forms: canonical, clinical, layperson, indirect or hedged, emotionally concerned, and deliberately under-specified. The first five express the same underlying concern and retain the same clinical information, whereas the under-specified form intentionally omits relevant details to test whether the model recognizes that clarification is needed. We evaluate a multilingual instruction model and QLoRA-adapted variants on concern classification, risk calibration, clarification behavior, parse compliance, and cross-form consistency. Results reveal that aggregate accuracy and consistency can conceal safety-relevant failures. A multilingual adaptation model reaches 0.994 under-triage in French and Arabic under language-asymmetric risk supervision. A controlled re-adaptation using source-derived, language-invariant risk labels reduces under-triage to 0.572 and 0.558, respectively. These findings show that robust multilingual healthcare evaluation requires explicit testing of register variation, uncertainty handling, and the provenance and invariance of adaptation labels.

Generally, most voice agents are cascaded systems, i.e., an ASR model transcribes the caller's audio, a language model reads the transcript and decides what to say and which backend tools to call, and a TTS model speaks the reply. Nearly all of the decision making happens in the language model, but existing evaluations measure it either too broadly or too narrowly. End-to-end voice benchmarks score the full pipeline, so recognition errors and model errors mix into a single number. LLM benchmarks isolate the model but they do not evaluate what makes real phone calls hard, such as transcription issues, caller's voice being split across messages and the requirement that replies follow the language and script specified. We introduce the Multi-Turn Voice Agent Benchmark (MTVA-Bench), which evaluates the language model on the same conditions it faces inside a cascaded system. The caller is played by an LLM following a set of rubrics and tool calls are answered by a mock backend which responds to the arguments the model actually sent. The benchmark contains 49 agents working across 490 reviewed scenarios and supports 7 languages. Scoring is a combination of deterministic checks on tool calls with two LLM judges, one that scores scenario specific rules and one that grades conversation quality without access to the task. Both judges must cite specific messages from the transcript. Task and conversation scores are weighted equally, since a call can complete its task and still go badly for the caller. In a seven-model study, six of the models select the correct tool within 6.4 points of one another, but their overall scores span 24.4 points. Most of the gap comes from argument values, action ordering, rule compliance, and what the model says around its tool calls.

People increasingly turn to AI chatbots for news and explanations of world events. But do they receive the same political answers when they ask in different languages? Here we show that the language of a question can change how the same AI systems assess the war in Ukraine. We ask GPT, Claude and Gemini to evaluate twenty statements about the war in 112 languages, collecting 67,200 responses. The balance between Russia-leaning and Ukraine-leaning responses differs across languages. When we group responses by countries' official languages, they follow a pattern resembling worldwide political divisions: relatively more Russia-leaning answers correspond to more favourable public views of Russia, less support for Ukraine in United Nations votes, and less aid to Ukraine. The broad pattern recurs across all three models and remains when individual statement pairs are removed. Our findings suggest a possible route through which information warfare may shape the text used to train AI models, which may in turn spread geopolitical biases.

We report the results of a Turing Test conducted in Finland in the Finnish language. Because languages and cultural contexts are unevenly represented in LLM training data, we expected the model (ChatGPT 5.2) to perform worse in a Finnish-language Turing Test than in previously studied English-language US contexts. We also present model-generated role prompting as a replicable technique for conducting comparative LLM-based Turing Tests designed to improve construct validity. Contrary to our expectations, the LLM passed the Finnish Turing Test. A prominent source of error was participants' reliance on linguistic cues, particularly colloquial Finnish, as markers of human authorship. We reframe the Turing Test from a test of intelligence to a comparative method for examining whether an AI system can display credible membership in a particular social world. Because its outcome reflects model capabilities, prompted identity, insider competence among human participants, and their AI literacy, the method provides a useful probe of the human-machine boundary across domains.