arXivDaily arXiv每日学术速递 周一至周五更新

高校专区

University of California, Los Angeles(加州大学洛杉矶分校)

2026-05-11 至 2026-05-11 共收录 4
2605.07723 2026-05-11 cs.DL cs.AI cs.CY physics.soc-ph

LLM hallucinations in the wild: Large-scale evidence from non-existent citations

在现实世界中大型语言模型的幻觉:来自不存在引用的大规模证据

Zhenyue Zhao, Yihe Wang, Toby Stuart, Mathijs De Vaan, Paul Ginsparg, Yian Yin

机构 * Department of Information Science, Cornell University(信息科学系,康奈尔大学) Department of Sociology, University of California Los Angeles(社会学系,加州大学洛杉矶分校) Department of Computer Science and Technology, Tsinghua University(计算机科学与技术系,清华大学) Haas School of Business, University of California Berkeley(哈斯商学院,加州大学伯克利分校)

AI总结 研究通过验证引用数据揭示LLM生成虚假引用的问题,发现2025年存在146932个虚假引用,且在AI应用快速发展的领域和语言特征显示AI辅助写作的论文中尤为严重,影响科学认可的公平性。

详情

展开后加载摘要…

URL PDF HTML 收藏
2605.07282 2026-05-11 cs.LG

The Convergence Gap: Instruction-Tuned Language Models Stabilize Later in the Forward Pass

收敛差距:指令微调语言模型在前向传递后期趋于稳定

Yifan Zhou

机构 * University of California, Los Angeles(加州大学洛杉矶分校)

AI总结 研究发现指令微调模型在前向传递后期更接近最终预测,通过分析六个预训练和指令微调检查点,发现后期MLP层对预测动态有显著影响。

详情

展开后加载摘要…

URL PDF HTML 收藏
2512.15567 2026-05-11 cs.AI cond-mat.mtrl-sci cs.LG physics.chem-ph

Evaluating Large Language Models in Scientific Discovery

评估大型语言模型在科学发现中的表现

Zhangde Song, Jieyu Lu, Yuanqi Du, Botao Yu, Thomas M. Pruyn, Yue Huang, Kehan Guo, Xiuzhe Luo, Yuanhao Qu, Yi Qu, Yinkai Wang, Haorui Wang, Jeff Guo, Jingru Gan, Parshin Shojaee, Di Luo, Andres M Bran, Gen Li, Qiyuan Zhao, Shao-Xiong Lennon Luo, Yuxuan Zhang, Xiang Zou, Wanru Zhao, Yifan F. Zhang, Wucheng Zhang, Shunan Zheng, Saiyang Zhang, Sartaaj Takrim Khan, Mahyar Rajabi-Kochi, Samantha Paradi-Maropakis, Tony Baltoiu, Fengyu Xie, Tianyang Chen, Kexin Huang, Weiliang Luo, Meijing Fang, Xin Yang, Lixue Cheng, Jiajun He, Soha Hassoun, Xiangliang Zhang, Wei Wang, Chandan K. Reddy, Chao Zhang, Zhiling Zheng, Mengdi Wang, Le Cong, Carla P. Gomes, Chang-Yu Hsieh, Aditya Nandy, Philippe Schwaller, Heather J. Kulik, Haojun Jia, Huan Sun, Seyed Mohamad Moosavi, Chenru Duan

机构 * Deep Principle(深原则) Department of Computer Science, Cornell University(计算机科学系,康奈尔大学) Department of Computer Science and Engineering, The Ohio State University(计算机科学与工程系,俄亥俄州立大学) Department of Chemical Engineering & Applied Chemistry, University of Toronto(化学工程与应用化学系,多伦多大学) Department of Computer Science and Engineering, University of Notre Dame(计算机科学与工程系,圣母大学) QuEra Computing Inc.(QuEra计算公司) Department of Pathology, Department of Genetics, Cancer Biology Program, Stanford University School of Medicine(病理学系、遗传学系、癌症生物学项目,斯坦福大学医学院) Harvard Law School(哈佛法学院) Department of Computer Science, Tufts University(计算机科学系,塔夫茨大学) School of Computational Science and Engineering, Georgia Institute of Technology(计算科学与工程学院,佐治亚理工学院) Department of Computer Science, University of California, Los Angeles(计算机科学系,加州大学洛杉矶分校) Department of Computer Science, Virginia Tech(计算机科学系,弗吉尼亚理工大学) Department of Physics, Tsinghua University(物理系,清华大学) Institute for Advanced Study, Tsinghua University(清华大学高级研究所) Laboratory of Artificial Chemical Intelligence, Ecole Polytechnique Federale de Lausanne(人工化学智能实验室,瑞士联邦理工学院)

AI总结 本文提出一个基于场景的基准测试,评估LLM在生物学、化学、材料科学和物理学中的科学发现能力,揭示了模型在科学发现任务中的性能差距和改进方向。

详情

展开后加载摘要…

URL PDF HTML 收藏
2506.13351 2026-05-11 cs.CL cs.AI cs.LG

Direct Reasoning Optimization: Token-Level Reasoning Reflectivity Meets Rubric Gates for Unverifiable Tasks

直接推理优化:基于推理反射的令牌级推理与评分门控用于不可验证任务

Yifei Xu, Tusher Chakraborty, Srinagesh Sharma, Leonardo Nunes, Swati Sharma, Kate Drakos Demopulos, Emre Kıcıman, Songwu Lu, Ranveer Chandra

机构 * Microsoft(微软公司) University of California, Los Angeles(加州大学洛杉矶分校)

AI总结 本文提出了一种约束强化学习框架,通过令牌级推理反射奖励和评分门控优化不可验证任务的推理质量与学习效率。

详情

展开后加载摘要…

URL PDF HTML 收藏