arXivDaily arXiv每日学术速递 周一至周五更新

高校专区

National University of Singapore(新加坡国立大学)

2026-04-24 至 2026-04-24 共收录 3
2604.02832 2026-04-24 q-fin.RM cs.LG

Transfer Learning for Loan Recovery Prediction under Distribution Shifts with Heterogeneous Feature Spaces

在存在分布偏移的情况下,用于贷款回收预测的迁移学习

Christopher Gerling, Hanqiu Peng, Ying Chen, Stefan Lessmann

机构 * National University of Singapore(新加坡国立大学)

AI总结 本文提出FT-MDN-Transformer模型,用于在异构特征集上进行贷款回收预测的迁移学习,通过蒙特卡洛模拟和真实数据集评估,证明其在数据稀缺和分布偏移下的有效性。

Comments 35 pages, 14 figures. Christopher Gerling had previously withdrawn his submission due to NDA restrictions, and that matter was resolved. We are authorized to publish the preprint now

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.05563 2026-04-24 cs.CV cs.SI

What's Left Unsaid? Detecting and Correcting Misleading Omissions in Multimodal News Previews

未被言说之物:检测和纠正多模态新闻预告中的误导性省略

Fanxiao Li, Jiaying Wu, Tingchao Fu, Dayang Li, Herun Wan, Wei Zhou, Min-Yen Kan

机构 * School of Information Science and Engineering, Yunnan University(云南大学信息科学与工程学院) National University of Singapore(新加坡国立大学) Xi’an Jiaotong University(西安交通大学) School of Engineering, Yunnan University(云南大学工程学院)

AI总结 本文提出MM-Misleading基准,通过多阶段流程评估开源LVLMs,发现省略性误导的盲区,并提出OMGuard结合细调和引导修正,提升检测与纠正效果,揭示局部叙事变化导致的误导。

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.06209 2026-04-24 cs.AI cs.CL

ReProbe: Efficient Test-Time Scaling of Multi-Step Reasoning by Probing Internal States of Large Language Models

ReProbe:通过探测大语言模型的内部状态实现多步骤推理的高效测试时间扩展

Jingwei Ni, Ekaterina Fadeeva, Tianyi Wu, Mubashara Akhtar, Jiaheng Zhang, Elliott Ash, Markus Leippold, Timothy Baldwin, See-Kiong Ng, Artem Shelmanov, Mrinmaya Sachan

机构 * ETH Zürich(苏黎世联邦理工学院) National University of Singapore(新加坡国立大学) MBZUAI(马斯喀特人工智能研究所) University of Zürich(苏黎世大学) The University of Melbourne(墨尔本大学)

AI总结 本文提出基于探测大语言模型内部状态的轻量级方法,用于多步骤推理验证,其性能超越了810倍大的Process Reward Models,为可扩展的测试时间扩展和更可 introspective 的大语言模型提供了新路径。

Comments ACL 2026 Main

详情

展开后加载摘要…

URL PDF HTML 收藏