Large language models (LLMs) struggle with cross-lingual knowledge transfer: they hallucinate when asked in one language about facts expressed in a different language during training. This work introduces a controlled setting to study the causes and dynamics of this phenomenon by training small Transformer models from scratch on synthetic multilingual datasets. We identify a learning phase wherein a model develops either separate or unified representations of the same facts across languages, and show that unification is essential for cross-lingual transfer. We also show that the degree of unification depends on mutual information between facts and training data language, and on how easy it is to extract that language. Based on these insights, we develop methods to modulate the level of cross-lingual transfer by manipulating data distribution and tokenization, and we introduce metrics and visualizations to formally characterize their effects on unification. Our work shows how controlled settings can shed light on pre-training dynamics and suggests new directions for improving cross-lingual transfer in LLMs. This behavior has been attributed to training and sampling noise, gaps in pretraining data (Xu et al., 2024), and misaligned incentives in post-training (Schul-man, 2023). However, these fail to explain cross-lingual factual errors: cases where models accurately answer questions when posed in the same language as the training data, yet hallucinate when prompted in a different (often lower-resource) language (Goldman et al., 2025). Failures of cross-lingual transfer exacerbate disadvantages faced by speakers of underrepresented languages, and increasing model scale does not solve the problem (Aggarwal et al., 2025; Qi et al., 2023). LLMs have been found to develop both a lingua franca for factual knowledge (typically based on English) and distinct language silos (Aggarwal et al., 2025; Lim et al., 2025b; Schut et al., 2025; Lim et al., 2025a, inter alia), and their hidden representations can be language-agnostic or language-specific depending on the layer (Wang et al., 2025). However the root cause of these phenomena is not understood, as most research on cross-lingual transfer analyzes models as static artifacts. Such analysis, while valuable, cannot explain how knowledge arises during training, and therefore cannot lead to effective pre-training interventions. While some have investigated the training dynamics of knowledge acquisition in multilingual LLMs (Zeng et al., 2025; Liu et al., 2025), their approach is non-interventional and does not establish a causal link between data properties and cross-lingual transfer.

From professional research to everyday planning, many tasks are bottlenecked by wide-scale information seeking, which is more repetitive than cognitively complex. With the rapid development of Large Language Models (LLMs), automated search agents powered by LLMs offer a promising solution to liberate humans from this tedious work. However, the capability of these agents to perform such "wide-context" collection reliably and completely remains largely unevaluated due to a lack of suitable benchmarks. To bridge this gap, we introduce WideSearch, a new benchmark engineered to evaluate agent reliability on these large-scale collection tasks. The benchmark features 200 manually curated questions (100 in English, 100 in Chinese) from over 15 diverse domains, grounded in real user queries. Each task requires agents to collect large-scale atomic information, which could be verified one by one objectively, and arrange it into a well-organized output. A rigorous five-stage quality control pipeline ensures the difficulty, completeness, and verifiability of the dataset. We benchmark over 10 state-of-the-art agentic search systems, including single-agent, multi-agent frameworks, and end-to-end commercial systems. Most systems achieve overall success rates near 0\%, with the best performer reaching just 5\%. However, given sufficient time, cross-validation by multiple human testers can achieve a near 100\% success rate. These results demonstrate that present search agents have critical deficiencies in large-scale information seeking, underscoring urgent areas for future research and development in agentic search. Our dataset, evaluation pipeline, and benchmark results have been publicly released at https://widesearch-seed.github.io/

Large Language Models (LLMs) are trained on a vast amount of procedural texts, but they do not directly observe real-world phenomena. In the context of cooking recipes, this poses a challenge, as intermediate states of ingredients are often omitted, making it difficult for models to track ingredient states and understand recipes accurately. In this paper, we apply state probing, a method for evaluating a language model's understanding of the world, to the domain of cooking. We propose a new task and dataset for evaluating how well LLMs can recognize intermediate ingredient states during cooking procedures. We first construct a new Japanese recipe dataset with clear and accurate annotations of ingredient state changes, collected from well-structured and controlled recipe texts. Using this dataset, we design three novel tasks to evaluate whether LLMs can track ingredient state transitions and identify ingredients present at intermediate steps. Our experiments with widely used LLMs, such as Llama3.1-70B and Qwen2.5-72B, show that learning ingredient state knowledge improves their understanding of cooking processes, achieving performance comparable to commercial LLMs. The dataset are publicly available at: https://huggingface.co/datasets/mashi6n/nhkrecipe-100-anno-1

Large Language Models (LLMs) can achieve strong performance on everyday coding tasks, but they can fail on complex tasks that require non-trivial reasoning about program semantics. Finding training examples to teach LLMs to solve these tasks can be challenging. In this work, we explore a method to synthetically generate code reasoning training data based on a semantic inequivalence game SInQ: a generator agent creates program variants that are semantically distinct, derived from a dataset of real-world programming tasks, while an evaluator agent has to identify input examples that cause the original programs and the generated variants to diverge in their behaviour, with the agents training each other semi-adversarially. We prove that this setup enables theoretically unlimited improvement through self-play in the limit of infinite computational resources. We evaluated our approach on multiple code generation and understanding benchmarks, including cross-language vulnerability detection (Lu et al., 2021), where our method improves vulnerability detection in C/C++ code despite being trained exclusively on Python code, and the challenging Python builtin identifier swap benchmark (Miceli-Barone et al., 2023), showing that whereas modern LLMs still struggle with this benchmark, our approach yields substantial improvements. We release the code needed to replicate the experiments, as well as the generated synthetic data, which can be used to fine-tune LLMs.

Large language models (LLMs) have demonstrated strong performance in sentence-level machine translation, but scaling to document-level translation remains challenging, particularly in modeling long-range dependencies and discourse phenomena across sentences and paragraphs. In this work, we propose a method to improve LLM-based long-document translation through targeted fine-tuning on high-quality document-level data, which we curate and introduce as DocBlocks. Our approach supports multiple translation paradigms, including direct document-to-document and chunk-level translation, by integrating instructions both with and without surrounding context. This enables models to better capture cross-sentence dependencies while maintaining strong sentence-level translation performance. Experimental results show that incorporating multiple translation paradigms improves document-level translation quality and inference speed compared to prompting and agent-based methods.

Multi-hop question answering faces substantial challenges due to data sparsity, which increases the likelihood of language models learning spurious patterns. To address this issue, prior research has focused on diversifying question generation through content planning and varied expression. However, these approaches often emphasize generating simple questions and neglect the integration of essential knowledge, such as relevant sentences within documents. This paper introduces the Knowledge Composition Sampling (KCS), an innovative framework designed to expand the diversity of generated multi-hop questions by sampling varied knowledge compositions within a given context. KCS models the knowledge composition selection as a sentence-level conditional prediction task and utilizes a probabilistic contrastive loss to predict the next most relevant piece of knowledge. During inference, we employ a stochastic decoding strategy to effectively balance accuracy and diversity. Compared to competitive baselines, our KCS improves the overall accuracy of knowledge composition selection by 3.9%, and its application for data augmentation yields improvements on HotpotQA and 2WikiMultihopQA datasets. Our code is available at: https://github.com/yangfanww/kcs.

This paper synthesizes the key findings from a full-day XP2025 workshop on "AI and Agile: From Frustration to Success", held in Brugg-Windisch, Switzerland. The workshop brought together over 30 interdisciplinary academic researchers and industry practitioners to tackle the concrete challenges and emerging opportunities at the intersection of Generative Artificial Intelligence (GenAI) and agile software development. Through structured, interactive breakout sessions, participants identified shared pain points like tool fragmentation, governance, data quality, and critical skills gaps in AI literacy and prompt engineering. These issues were further analyzed, revealing underlying causes and cross-cutting concerns. The workshop concluded by collaboratively co-creating a multi-thematic research roadmap, articulating both short-term, implementable actions and visionary, long-term research directions. This cohesive agenda aims to guide future investigation and drive the responsible, human-centered integration of GenAI into agile practices.

In the dynamic landscape of large-scale web search, Query-Driven Text Summarization (QDTS) aims to generate concise and informative summaries from textual documents based on a given query, which is essential for improving user engagement and facilitating rapid decision-making. Traditional extractive summarization models, based primarily on ranking candidate summary segments, have been the dominant approach in industrial applications. However, these approaches suffer from two key limitations: 1) The multi-stage pipeline often introduces cumulative information loss and architectural bottlenecks due to its weakest component; 2) Traditional models lack sufficient semantic understanding of both user queries and documents, particularly when dealing with complex search intents. In this study, we propose a novel framework to pioneer the application of generative models to address real-time QDTS in industrial web search. Our approach integrates large model distillation, supervised fine-tuning, direct preference optimization, and lookahead decoding to transform a lightweight model with only 0.1B parameters into a domain-specialized QDTS expert. Evaluated on multiple industry-relevant metrics, our model outperforms the production baseline and achieves a new state of the art. Furthermore, it demonstrates excellent deployment efficiency, requiring only 334 NVIDIA L20 GPUs to handle \textasciitilde50,000 queries per second under 55~ms average latency per query.

The widespread success of pre-trained language models has established a new training paradigm, where a global PLM is fine-tuned using task-specific data from local clients. The local data are highly different from each other and can not capture the global distribution of the whole data in real world. To address the challenges of non-IID data in real environments, privacy-preserving federated distillation has been proposed and highly investigated. However, previous experimental non-IID scenarios are primarily identified with the label (output) diversity, without considering the diversity of language domains (input) that is crucial in natural language processing. In this paper, we introduce a comprehensive set of multi-domain non-IID scenarios and propose a unified benchmarking framework that includes diverse data. The benchmark can be used to evaluate the federated learning framework in a real environment. To this end, we propose an Adaptive Federated Distillation (AdaFD) framework designed to address multi-domain non-IID challenges in both homogeneous and heterogeneous settings. Experimental results demonstrate that our models capture the diversity of local clients and achieve better performance compared to the existing works. The code for this paper is available at: https://github.com/jiahaoxiao1228/AdaFD.

This is an overview of the thirteenth edition of the BioASQ challenge in the context of the Conference and Labs of the Evaluation Forum (CLEF) 2025. BioASQ is a series of international challenges promoting advances in large-scale biomedical semantic indexing and question answering. This year, BioASQ consisted of new editions of the two established tasks, b and Synergy, and four new tasks: a) Task MultiClinSum on multilingual clinical summarization. b) Task BioNNE-L on nested named entity linking in Russian and English. c) Task ELCardioCC on clinical coding in cardiology. d) Task GutBrainIE on gut-brain interplay information extraction. In this edition of BioASQ, 83 competing teams participated with more than 1000 distinct submissions in total for the six different shared tasks of the challenge. Similar to previous editions, several participating systems achieved competitive performance, indicating the continuous advancement of the state-of-the-art in the field.