Over the past few years, Large Language Models (LLMs) have demonstrated remarkable capabilities across a wide range of natural language processing tasks (Xie et al., 2023; Zhang et al., 2023; Gao et al., 2024; OpenAI et al., 2024). However, they are still prone to generating hallucinations--content that is not supported by the input context or the underlying knowledge sources (Zhu et al., 2024; Kalai et al., 2025; Huang et al., 2025). Hallucinations pose critical risks in downstream applications such as summarization and long-form question answering, where reliability and factual consistency with respect to the input context are paramount. Hence, the ability to detect hallucinations is crucial for successful real-world deployment of LLMs. Most existing research works focus on binary hallucination detection problem, where the goal is to determine if the model output contains hallucinations or not (Yang et al., 2024a,b; Tang et al., 2024; Ravi et al., 2024; Ji et al., 2024; Chuang et al., 2024). While useful, this formulation is limited: in many real-world applications, one often needs to know which specific spans in the model output are hallucinated in order to assess the reliability of the generated content. This motivates the problem of hallucination span detection, where the goal is to precisely locate unsupported content in the model output (Wu et al., 2023; Ogasa and Arase, 2025). Recently, reasoning--the process of systematically arriving at conclusions by generating and utilizing intermediate steps--has been shown to significantly enhance the capabilities of LLMs in solving complex tasks such as mathematics (Shao et al., 2024; Yu et al., 2025) and coding (Liu and Zhang, 2025; Chen et al., 2025). Hallucination span detection is also a complex multi-step decision making process as it requires carefully analyzing the model output to extract all the stated facts and verifying whether each of these facts is fully supported by the input context, and could benefit significantly from a learned reasoning process.

Toxic speech on online platforms is a growing concern, impacting user experience and online safety. While text-based toxicity detection is well-studied, audio-based approaches remain underexplored, especially for low-resource languages like Vietnamese. This paper introduces ViToSA (Vietnamese Toxic Spans Audio), the first dataset for toxic spans detection in Vietnamese speech, comprising 11,000 audio samples (25 hours) with accurate human-annotated transcripts. We propose a pipeline that combines ASR and toxic spans detection for fine-grained identification of toxic content. Our experiments show that fine-tuning ASR models on ViToSA significantly reduces WER when transcribing toxic speech, while the text-based toxic spans detection (TSD) models outperform existing baselines. These findings establish a novel benchmark for Vietnamese audio-based toxic spans detection, paving the way for future research in speech content moderation.

This paper presents one of the top-performing solutions to the UNLP 2025 Shared Task on Detecting Manipulation in Social Media. The task focuses on detecting and classifying rhetorical and stylistic manipulation techniques used to influence Ukrainian Telegram users. For the classification subtask, we fine-tuned the Gemma 2 language model with LoRA adapters and applied a second-level classifier leveraging meta-features and threshold optimization. For span detection, we employed an XLM-RoBERTa model trained for multi-target, including token binary classification. Our approach achieved 2nd place in classification and 3rd place in span detection.

Despite the recent success of two-stage prototypical networks in few-shot named entity recognition (NER), challenges such as over/under-detected false spans in the span detection stage and unaligned entity prototypes in the type classification stage persist. Additionally, LLMs have not proven to be effective few-shot information extractors in general. In this paper, we propose an approach called Boundary-Aware LLMs for Few-Shot Named Entity Recognition to address these issues. We introduce a boundary-aware contrastive learning strategy to enhance the LLM's ability to perceive entity boundaries for generalized entity spans. Additionally, we utilize LoRAHub to align information from the target domain to the source domain, thereby enhancing adaptive cross-domain classification capabilities. Extensive experiments across various benchmarks demonstrate that our framework outperforms prior methods, validating its effectiveness. In particular, the proposed strategies demonstrate effectiveness across a range of LLM architectures. The code and data are released on https://github.com/UESTC-GQJ/BANER.

Bangla is the seventh most spoken language by a total number of speakers in the world, and yet the development of an automated grammar checker in this language is an understudied problem. Bangla grammatical error detection is a task of detecting sub-strings of a Bangla text that contain grammatical, punctuation, or spelling errors, which is crucial for developing an automated Bangla typing assistant. Our approach involves breaking down the task as a token classification problem and utilizing state-of-the-art transformer-based models. Finally, we combine the output of these models and apply rule-based post-processing to generate a more reliable and comprehensive result. Our system is evaluated on a dataset consisting of over 25,000 texts from various sources. Our best model achieves a Levenshtein distance score of 1.04. Finally, we provide a detailed analysis of different components of our system.

Medical decisions directly impact individuals' health and well-being. Extracting decision spans from clinical notes plays a crucial role in understanding medical decision-making processes. In this paper, we develop a new dataset called "MedDec", which contains clinical notes of eleven different phenotypes (diseases) annotated by ten types of medical decisions. We introduce the task of medical decision extraction, aiming to jointly extract and classify different types of medical decisions within clinical notes. We provide a comprehensive analysis of the dataset, develop a span detection model as a baseline for this task, evaluate recent span detection approaches, and employ a few metrics to measure the complexity of data samples. Our findings shed light on the complexities inherent in clinical decision extraction and enable future work in this area of research. The dataset and code are available through https://github.com/CLU-UML/MedDec.

Abuse in its various forms, including physical, psychological, verbal, sexual, financial, and cultural, has a negative impact on mental health. However, there are limited studies on applying natural language processing (NLP) in this field in Vietnam. Therefore, we aim to contribute by building a human-annotated Vietnamese dataset for detecting abusive content in Vietnamese narrative texts. We sourced these texts from VnExpress, Vietnam's popular online newspaper, where readers often share stories containing abusive content. Identifying and categorizing abusive spans in these texts posed significant challenges during dataset creation, but it also motivated our research. We experimented with lightweight baseline models by freezing PhoBERT and XLM-RoBERTa and using their hidden states in a BiLSTM to assess the complexity of the dataset. According to our experimental results, PhoBERT outperforms other models in both labeled and unlabeled abusive span detection tasks. These results indicate that it has the potential for future improvements.

We introduce the submissions of the NJUNLP team to the WMT 2023 Quality Estimation (QE) shared task. Our team submitted predictions for the English-German language pair on all two sub-tasks: (i) sentence- and word-level quality prediction; and (ii) fine-grained error span detection. This year, we further explore pseudo data methods for QE based on NJUQE framework (https://github.com/NJUNLP/njuqe). We generate pseudo MQM data using parallel data from the WMT translation task. We pre-train the XLMR large model on pseudo QE data, then fine-tune it on real QE data. At both stages, we jointly learn sentence-level scores and word-level tags. Empirically, we conduct experiments to find the key hyper-parameters that improve the performance. Technically, we propose a simple method that covert the word-level outputs to fine-grained error span results. Overall, our models achieved the best results in English-German for both word-level and fine-grained error span detection sub-tasks by a considerable margin.

As hate speech continues to proliferate on the web, it is becoming increasingly important to develop computational methods to mitigate it. Reactively, using black-box models to identify hateful content can perplex users as to why their posts were automatically flagged as hateful. On the other hand, proactive mitigation can be achieved by suggesting rephrasing before a post is made public. However, both mitigation techniques require information about which part of a post contains the hateful aspect, i.e., what spans within a text are responsible for conveying hate. Better detection of such spans can significantly reduce explicitly hateful content on the web. To further contribute to this research area, we organized HateNorm at HASOC-FIRE 2023, focusing on explicit span detection in English Tweets. A total of 12 teams participated in the competition, with the highest macro-F1 observed at 0.58.

We present the joint contribution of Unbabel and Instituto Superior T\'ecnico to the WMT 2023 Shared Task on Quality Estimation (QE). Our team participated on all tasks: sentence- and word-level quality prediction (task 1) and fine-grained error span detection (task 2). For all tasks, we build on the COMETKIWI-22 model (Rei et al., 2022b). Our multilingual approaches are ranked first for all tasks, reaching state-of-the-art performance for quality estimation at word-, span- and sentence-level granularity. Compared to the previous state-of-the-art COMETKIWI-22, we show large improvements in correlation with human judgements (up to 10 Spearman points). Moreover, we surpass the second-best multilingual submission to the shared-task with up to 3.8 absolute points.