Despite multilingual pretraining, large language models often struggle with non-English tasks, particularly in language control, the ability to respond in the intended language. We identify and characterize two key failure modes: the multilingual transfer bottleneck (correct language, incorrect task response) and the language consistency bottleneck (correct task response, wrong language). To systematically surface these issues, we design a four-scenario evaluation protocol spanning MMLU, MGSM, and XQuAD benchmarks. To probe these issues with interpretability, we extend logit lens analysis to track language probabilities layer by layer and compute cross-lingual semantic similarity of hidden states. The results reveal a three-phase internal structure: early layers align inputs into a shared semantic space, middle layers perform task reasoning, and late layers drive language-specific generation. Guided by these insights, we introduce selective fine-tuning of only the final layers responsible for language control. On Qwen-3-32B and Bloom-7.1B, this method achieves over 98 percent language consistency across six languages while fine-tuning only 3-5 percent of parameters, without sacrificing task accuracy. Importantly, this result is nearly identical to that of full-scope fine-tuning (for example, above 98 percent language consistency for both methods across all prompt scenarios) but uses a fraction of the computational resources. To the best of our knowledge, this is the first approach to leverage layer-localization of language control for efficient multilingual adaptation.
The Impact of Large Language Models in Academia: from Writing to Speaking
International School for Advanced Studies (SISSA)（国际高等研究院（SISSA））；Huazhong Univerisity of Science and Technology（华中科技大学）；University of Washington（华盛顿大学）
Let LLMs Speak Embedding Languages: Generative Text Embeddings via Iterative Contrastive Refinement
Department of Computer Science and Information Engineering, National Taiwan University（国家台湾大学计算机科学与信息工程系）；National Taiwan University AI Center of Research Excellence（国家台湾大学人工智能研究中心）
LaERC-S: Improving LLM-based Emotion Recognition in Conversation with Speaker Characteristics
School of Computer Science and Technology, Harbin Institute of Technology（哈尔滨工业大学计算机科学与技术学院）；School of Computer Science and Technology, Soochow University（苏州大学计算机科学与技术学院）；School of Computer Science and Technology, Harbin Institute of Technology, Shenzhen（哈尔滨工业大学（深圳）计算机科学与技术学院）
Comparing Moral Values in Western English-speaking societies and LLMs with Word Associations
School of Computing and Information Systems, The University of Melbourne（墨尔本大学计算与信息系统学院）；Complex Human Data Hub, The University of Melbourne（墨尔本大学复杂人类数据中心）
为避免直接提示的局限性，本研究提出利用词汇联想作为底层表征，通过在人类与LLM联想图中传播Moral Foundation Theory的种子词，系统比较了西方英语社会与LLM在道德价值观上的差异。
Atom-anchored LLMs speak Chemistry: A Retrosynthesis Demonstration
Machine Learning Research（机器学习研究）；Pfizer Research and Development（辉瑞研发）；Leiden Institute of Advanced Computer Science（莱顿高级计算机科学研究所）；Leiden University（莱顿大学）；Leiden Academic Centre for Drug Research（莱顿药物研究中心）；Leiden Institute of Chemistry（莱顿化学研究所）
Uncertainty quantification in fine-tuned LLMs using LoRA ensembles
Stockholm University Department of Physics（斯德哥尔摩大学物理系）；Department of Mathematical Sciences（数学科学系）；Chalmers university of technology & University of Gothenburg（楚德勒斯技术大学及哥德堡大学）；VERSES Research Lab（VERSES研究实验室）
Jailbreak Instruction-Tuned LLMs via end-of-sentence MLP Re-weighting
Peking University（北京大学）；School of Science（理学院）；Westlake University（西湖大学）；Beijing International Center for Mathematical Research（北京国际数学研究中心）；Center for Machine Learning Research（机器学习研究中心）
60 Data Points are Sufficient to Fine-Tune LLMs for Question-Answering
School of Computer Science, Fudan University（复旦大学计算机科学学院）；Institute of Modern Languages and Linguistics, Fudan University（复旦大学现代语言与语言学研究所）；Lenovo Research（联想研究院）
AlignedCoT: Prompting Large Language Models via Native-Speaking Demonstrations
The Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州））；City University of Hong Kong（香港城市大学）；Shenzhen Campus of Sun Yat-sen University（中山大学深圳校区）；The Hong Kong University of Science and Technology（香港科技大学）；University of California, Merced（加州大学默塞德分校）；Chongqing University（重庆大学）；MBZUAI（穆罕默德·本·扎耶德人工智能大学）
Summarize Before You Speak with ARACH: A Training-Free Inference-Time Plug-In for Enhancing LLMs via Global Attention Reallocation
Meakins-Christe Laboratories, Research Institute of McGill University Health Centre（麦金-克里斯特实验室，麦吉尔大学健康中心研究 institute）；Department of Medicine, Division of Experimental Medicine, McGill University（医学系，实验医学 division，麦吉尔大学）；The College of Computer Science and Technology, Zhejiang Gongshang University（浙江工商大学计算机科学与技术学院）；Zhejiang Key Laboratory of Big Data and Future E-Commerce Technology（浙江大数据与未来电子商务技术重点实验室）；Quantitative Life Sciences, McGill University（定量生命科学，麦吉尔大学）；School of Computer Science, McGill University（计算机科学学院，麦吉尔大学）；Mila-Quebec AI Institute（蒙特利尔 AI 机构，魁北克）
Whose Morality Do They Speak? Unraveling Cultural Bias in Multilingual Language Models
Research Center Trustworthy Data Science and Security of the University Alliance Ruhr（鲁尔大学联盟可信数据科学与安全研究中心）；Faculty of Informatics, Technical University Dortmund（多特蒙德工业大学信息学院）
