Reliable financial reasoning requires knowing not only how to answer, but also when an answer cannot be justified. In real financial practice, problems often rely on implicit assumptions that are taken for granted rather than stated explicitly, causing problems to appear solvable while lacking enough information for a definite answer. We introduce RealFin, a bilingual benchmark that evaluates financial reasoning by systematically removing essential premises from exam-style questions while keeping them linguistically plausible. Based on this, we evaluate models under three formulations that test answering, recognizing missing information, and rejecting unjustified options, and find consistent performance drops when key conditions are absent. General-purpose models tend to over-commit and guess, while most finance-specialized models fail to clearly identify missing premises. These results highlight a critical gap in current evaluations and show that reliable financial models must know when a question should not be answered. The dataset and code are available athttps://github.com/insait-institute/RealFin.

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.

Multi-step symbolic reasoning is essential for robust financial analysis; yet, current benchmarks largely overlook this capability. Existing datasets such as FinQA and ConvFinQA emphasize final numerical answers while neglecting the intermediate reasoning steps required for transparency and verification. To address this gap, we introduce FinChain, the first benchmark specifically designed for verifiable Chain-of-Thought evaluation in finance. FinChain spans 58 topics across 12 financial domains, each represented by parameterized symbolic templates with executable Python code that enable fully machine-verifiable reasoning and scalable, contamination-free data generation.To assess reasoning capacity, we propose ChainEval, a dynamic alignment measure that jointly evaluates both the final-answer correctness and the step-level reasoning consistency. Our evaluation of 26 leading LLMs reveals that even frontier LLMs exhibit clear limitations in symbolic financial reasoning, while domain-adapted and math-enhanced fine-tuned models can substantially narrow this gap.Overall, FinChain exposes persistent weaknesses in multi-step financial reasoning and provides a foundation for developing trustworthy, interpretable, and verifiable financial AI. This project is available at https://github.com/mbzuai-nlp/finchain.git.

There is a significant gap in evaluating cultural reasoning in LLMs using conversational datasets that capture culturally rich and dialectal contexts. Most Arabic benchmarks focus on short text snippets in Modern Standard Arabic (MSA), overlooking the cultural nuances that naturally arise in dialogues. To address this gap, we introduce ArabCulture-Dialogue, a culturally grounded conversational dataset covering 13 Arabic-speaking countries, in both MSA and each country’s respective dialect, spanning 12 daily-life topics and 54 fine-grained subtopics. We utilize the dataset to form three benchmarking tasks: (i) multiple-choice cultural reasoning, (ii) machine translation between MSA and dialects, and (iii) dialect-steering generation. Our experiments indicate that the performance gap between MSA and Arabic dialects still exists, whereby the models perform worse on all three tasks in the dialectal setup, compared to the MSA one.

Large language models (LLMs) have been widely applied across various domains of finance. Since their training data are largely derived from human-authored corpora, LLMs may inherit a range of human biases. Behavioral biases can lead to instability and uncertainty in decision-making, particularly when processing financial information. However, existing research on LLM bias has mainly focused on direct questioning or simplified, general-purpose settings, with limited consideration of the complex real-world financial environments and high-risk, context-sensitive, multilingual financial misinformation detection tasks (MFMD). In this work, we propose MFMDScen, a comprehensive benchmark for evaluating behavioral biases of LLMs in MFMD across diverse economic scenarios. In collaboration with financial experts, we construct three types of complex financial scenarios: (i) role- and personality-based, (ii) role- and region-based, and (iii) role-based scenarios incorporating ethnicity and religious beliefs. We further develop a multilingual financial misinformation dataset covering English, Chinese, Greek, and Bengali. By integrating these scenarios with misinformation claims, MFMDScen enables a systematic evaluation of 22 mainstream LLMs. Our findings reveal that pronounced behavioral biases persist across both commercial and open-source models. This project is available at https://github.com/lzw108/FMD.

Financial question answering (QA) over long corporate filings requires evidence to satisfy strict constraints on entities, financial metrics, fiscal periods, and numeric values. However, existing LLM-based rerankers primarily optimize semantic relevance, leading to unstable rankings and opaque decisions on long documents. We propose FINCARDS, a structured reranking framework that reframes financial evidence selection as constraint satisfaction under a finance-aware schema. FINCARDS represents filing chunks and questions using aligned schema fields (entities, metrics, periods, and numeric spans), enabling deterministic field-level matching. Evidence is selected via a multi-stage tournament reranking with stability-aware aggregation, producing auditable decision traces. Across two corporate filing QA benchmarks, FINCARDS substantially improves early-rank retrieval over both lexical and LLM-based reranking baselines, while reducing ranking variance, without requiring model fine-tuning or unpredictable inference budgets. Our code is available at https://github.com/XanderZhou2022/FINCARDS.

English financial NLP has progressed rapidly through benchmarks for sentiment, document understanding, and financial question answering, while Arabic financial NLP remains comparatively under-explored despite strong practical demand for trustworthy finance and Islamic-finance assistants. We introduce SAHM, a document-grounded benchmark and instruction-tuning dataset for Arabic financial NLP and Shari’ah-compliant reasoning. SAHM contains 14,380 expert-verified instances spanning seven tasks: AAOIFI standards QA, fatwa-based QA/MCQ, accounting and business exams, financial sentiment analysis, extractive summarization, and event–cause reasoning, curated from authentic regulatory, juristic, and corporate sources. We evaluate 19 strong open and proprietary LLMs using task-specific metrics and rubric-based scoring for open-ended outputs, and find that Arabic fluency does not reliably translate to evidence-grounded financial reasoning: models are substantially stronger on recognition-style tasks than on generation and causal reasoning, with the largest gaps on event–cause reasoning. We release the benchmark, evaluation framework, and an instruction-tuned model to support future research on trustworthy Arabic financial NLP.

The rapid advancement of vision-language models (VLMs) has brought a lot of attention to their safety alignment. However, existing methods have primarily focused on model undersafety, where the model responds to hazardous queries, while neglecting oversafety, where the model refuses to answer safe queries. In this paper, we introduce the concept of safety calibration, which systematically addresses both undersafety and oversafety. Specifically, we present VSCBench, a novel dataset of 3,600 image-text pairs that are visually or textually similar but differ in terms of safety, which is designed to evaluate safety calibration across image-centric and text-centric scenarios. Based on our benchmark, we evaluate safety calibration across eleven widely used VLMs. Our extensive experiments revealed major issues with both undersafety and oversafety. We further investigated four approaches to improve the model’s safety calibration. We found that even though some methods effectively calibrated the models’ safety problems, these methods also lead to the degradation of models’ utility. This trade-off underscores the urgent need for advanced calibration methods, and our benchmark provides a valuable tool for evaluating future approaches.

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.

Despite having a population of twenty million, Kazakhstan’s culture and language remain underrepresented in the field of natural language processing. Although large language models (LLMs) continue to advance worldwide, progress in Kazakh language has been limited, as seen in the scarcity of dedicated models and benchmark evaluations. To address this gap, we introduce KazMMLU, the first MMLU-style dataset specifically designed for Kazakh language. KazMMLU comprises 23,000 questions that cover various educational levels, including STEM, humanities, and social sciences, sourced from authentic educational materials and manually validated by native speakers and educators. The dataset includes 10,969 Kazakh questions and 12,031 Russian questions, reflecting Kazakhstan’s bilingual education system and rich local context. Our evaluation of several state-of-the-art multilingual models (Llama3.1, Qwen-2.5, GPT-4, and DeepSeek V3) demonstrates substantial room for improvement, as even the best-performing models struggle to achieve competitive performance in Kazakh and Russian. These findings highlight significant performance gaps compared to high-resource languages. We hope that our dataset will enable further research and development of Kazakh-centric LLMs.

We introduce a novel multilingual and hierarchical corpus annotated for entity framing and role portrayal in news articles. The dataset uses a unique taxonomy inspired by storytelling elements, comprising 22 fine-grained roles, or archetypes, nested within three main categories: protagonist, antagonist, and innocent. Each archetype is carefully defined, capturing nuanced portrayals of entities such as guardian, martyr, and underdog for protagonists; tyrant, deceiver, and bigot for antagonists; and victim, scapegoat, and exploited for innocents. The dataset includes 1,378 recent news articles in five languages (Bulgarian, English, Hindi, European Portuguese, and Russian) focusing on two critical domains of global significance: the Ukraine-Russia War and Climate Change. Over 5,800 entity mentions have been annotated with role labels. This dataset serves as a valuable resource for research into role portrayal and has broader implications for news analysis. We describe the characteristics of the dataset and the annotation process, and we report evaluation results on fine-tuned state-of-the-art multilingual transformers and hierarchical zero-shot learning using LLMs at the level of a document, a paragraph, and a sentence.

We introduce SemEval-2025 Task 10 on Multilingual Characterization and Extraction of Narratives from Online News, which focuses on the identification and analysis of narratives in online news media. The task is structured into three subtasks: (1) Entity Framing, to identify the roles that relevant entities play within narratives, (2) Narrative Classification, to assign documents fine-grained narratives according to a given, topic-specific taxonomy of narrative labels, and (3) Narrative Extraction, to provide a justification for the dominant narrative of the document. To this end, we analyze news articles across two critical domains, Ukraine-Russia War and Climate Change, in five languages: Bulgarian, English, Hindi, Portuguese, and Russian. This task introduces a novel multilingual and multifaceted framework for studying how online news media construct and disseminate manipulative narratives. By addressing these challenges, our work contributes to the broader effort of detecting, understanding, and mitigating the spread of propaganda and disinformation. The task attracted a lot of interest: 310 teams registered, with 66 submitting official results on the test set.

We describe our system for SemEval-2025 Task 10 Subtask 1 on coarse-grained entity framing in Hindi news, exploring two complementary strategies. First, we experiment with LLM prompting using GPT-4o, comparing hierarchical multi-step prompting with native single-step prompting for both main and fine-grained role prediction. Second, we conduct an extensive study on fine-tuning XLM-R, analyzing different context granularities (full article, paragraph, or sentence-level entity mentions), monolingual vs. multilingual settings, and main vs. fine-grained role labels. Our best system, trained on fine-grained role annotations across languages using sentence-level context, achieved 43.99% exact match, 56.56 % precision, 47.38% recall, and 51.57% F1-score. Notably, our system set a new state-of-the-art for main role prediction on Hindi news, achieving 78.48 % accuracy - outperforming the next best model at 76.90%, as per the official leaderboard. Our findings highlight effective strategies for entity framing in multilingual and low-resource settings.

Fact-checking long-form text is challenging, and it is therefore common practice to break it down into multiple atomic claims. The typical approach to fact-checking these atomic claims involves retrieving a fixed number of pieces of evidence, followed by a verification step. However, this method is usually not cost-effective, as it underutilizes the verification model’s internal knowledge of the claim and fails to replicate the iterative reasoning process in human search strategies. To address these limitations, we propose FIRE, a novel agent-based framework that integrates evidence retrieval and claim verification in an iterative manner. Specifically, FIRE employs a unified mechanism to decide whether to provide a final answer or generate a subsequent search query, based on its confidence in the current judgment. We compare FIRE with other strong fact-checking frameworks and find that it achieves slightly better performance while reducing large language model (LLM) costs by an average of 7.6 times and search costs by 16.5 times. These results indicate that FIRE holds promise for application in large-scale fact-checking operations.

LLMs have the tendency to hallucinate, i.e., to sporadically generate false or fabricated information, and users generally lack the tools to detect when this happens. Uncertainty quantification (UQ) provides a framework for assessing the reliability of model outputs, aiding in the identification of potential hallucinations. In this work, we introduce pre-trained UQ heads: supervised auxiliary modules for LLMs that substantially enhance their ability to capture uncertainty compared to unsupervised UQ methods. Their strong performance stems from the transformer architecture in their design, in the form of informative features derived from LLM attention maps and logits. Our experiments show that these heads are highly robust and achieve state-of-the-art performance in claim-level hallucination detection across both in-domain and out-of-domain prompts. Moreover, these modules demonstrate strong generalization to languages they were not explicitly trained on. We pre-train a collection of UQ heads for popular LLM series, including Mistral, Llama, and Gemma. We publicly release both the code and the pre-trained heads.

The ease of access to large language models (LLMs) has enabled a widespread of machine-generated texts, and now it is often hard to tell whether a piece of text was human-written or machine-generated. This raises concerns about potential misuse, particularly within educational and academic domains. Thus, it is important to develop practical systems that can automate the process. Here, we present one such system, LLM-DetectAIve, designed for fine-grained detection. Unlike most previous work on machine-generated text detection, which focused on binary classification, LLM-DetectAIve supports four categories: (i) human-written, (ii) machine-generated, (iii) machine-written, then machine-humanized, and (iv) human-written, then machine-polished. Category (iii) aims to detect attempts to obfuscate the fact that a text was machine-generated, while category (iv) looks for cases where the LLM was used to polish a human-written text, which is typically acceptable in academic writing, but not in education. Our experiments show that LLM-DetectAIve can effectively identify the above four categories, which makes it a potentially useful tool in education, academia, and other domains.LLM-DetectAIve is publicly accessible at https://github.com/mbzuai-nlp/LLM-DetectAIve. The video describing our system is available at https://youtu.be/E8eT_bE7k8c.

Numerous evaluation metrics have been developed for natural language generation tasks, but their effectiveness in evaluating stories is limited as they are not specifically tailored to assess intricate aspects of storytelling, such as fluency and interestingness. In this paper, we introduce DeltaScore, a novel methodology that uses perturbation techniques for the evaluation of nuanced story aspects. We posit that the extent to which a story excels in a specific aspect (e.g., fluency) correlates with the magnitude of its susceptibility to particular perturbations (e.g., the introduction of typos). Given this, we measure the quality of an aspect by calculating the likelihood difference between pre- and post-perturbation states using pre-trained language models. We compare DeltaScore with existing metrics on storytelling datasets from two domains in five fine-grained story aspects: fluency, coherence, relatedness, logicality, and interestingness. DeltaScore demonstrates strong performance, revealing a surprising finding that one specific perturbation proves highly effective in capturing multiple aspects. Source code is available on our GitHub repository.

To enhance the quality of generated stories, recent story generation models have been investigating the utilization of higher-level attributes like plots or commonsense knowledge. The application of prompt-based learning with large language models (LLMs), exemplified by GPT-3, has exhibited remarkable performance in diverse natural language processing (NLP) tasks. This paper conducts a comprehensive investigation, utilizing both automatic and human evaluation, to compare the story generation capacity of LLMs with recent models across three datasets with variations in style, register, and length of stories. The results demonstrate that LLMs generate stories of significantly higher quality compared to other story generation models. Moreover, they exhibit a level of performance that competes with human authors, albeit with the preliminary observation that they tend to replicate real stories in situations involving world knowledge, resembling a form of plagiarism.

GPT-2 has been frequently adapted in story generation models as it provides powerful generative capability. However, it still fails to generate consistent stories and lacks diversity. Current story generation models leverage additional information such as plots or commonsense into GPT-2 to guide the generation process. These approaches focus on improving generation quality of stories while our work look at both quality and diversity. We explore combining BERT and GPT-2 to build a variational autoencoder (VAE), and extend it by adding additional objectives to learn global features such as story topic and discourse relations. Our evaluations show our enhanced VAE can provide better quality and diversity trade off, generate less repetitive story content and learn a more informative latent variable.

In this paper, we adapt Deep-speare, a joint neural network model for English sonnets, to Chinese poetry. We illustrate characteristics of Chinese quatrain and explain our architecture as well as training and generation procedure, which differs from Shakespeare sonnets in several aspects. We analyse the generated poetry and find that model works well for Chinese poetry, as it can: (1) generate coherent 4-line quatrains of different topics; and (2) capture rhyme automatically (to a certain extent).