arXivDaily arXiv每日学术速递 周一至周五更新

期刊&会议

International Conference on Learning Representations · 会议 · Machine Learning

2026-07-03 至 2026-07-03 共收录 3
2603.05256 2026-07-03 cs.CV 版本更新

Wiki-R1: Incentivizing Multimodal Reasoning for Knowledge-based VQA via Data and Sampling Curriculum

Wiki-R1: 通过数据和采样课程激励基于知识的多模态推理用于VQA

Shan Ning, Longtian Qiu, Xuming He

机构 * ShanghaiTech University(上海科技大学) Shanghai Engineering Research Center of Intelligent Vision and Imaging(上海智能视觉与成像工程研究中心) Lingang Laboratory(临港实验室)

AI总结 提出Wiki-R1框架,通过可控课程数据生成和课程采样策略,结合强化学习激励MLLMs在KB-VQA中的推理能力,在Encyclopedic VQA和InfoSeek上取得新SOTA。

Comments Accepted by ICLR 26, code and weights are publicly available

详情

展开后加载摘要…

URL PDF HTML 收藏
2408.01139 2026-07-03 cs.AI cs.CV 版本更新

Interpreting Global Perturbation Robustness of Image Models using Axiomatic Spectral Importance Decomposition

使用公理谱重要性分解解释图像模型的全局扰动鲁棒性

Róisín Luo, James McDermott, Colm O'Riordan

机构 * SFI Centre for Research Training in Artificial Intelligence(SFI人工智能研究培训中心) School of Computer Science, University of Galway(Galway大学计算机科学学院)

AI总结 提出一种模型无关的全局可解释性方法I-ASIDE,基于Shapley值公理量化鲁棒与非鲁棒特征的预测能力,揭示图像模型对数据损坏和对抗攻击等扰动的鲁棒性机制。

Comments Accepted by Transactions on Machine Learning Research (TMLR 2024)

Journal ref Transactions on Machine Learning Research (TMLR), 2024; Presented at The Thirteenth International Conference on Learning Representations (ICLR 2025), Singapore

详情

展开后加载摘要…

URL PDF HTML 收藏
2504.02327 2026-07-03 cs.CL 版本更新

LearNAT: Learning NL2SQL with AST-guided Task Decomposition for Large Language Models

LearNAT: 基于AST引导任务分解的NL2SQL大语言模型学习

Weibin Liao, Xin Gao, Tianyu Jia, Rihong Qiu, Yifan Zhu, Yang Lin, Xinyu Ma, Junfeng Zhao, Yasha Wang

机构 * School of Computer Science, Peking University(北京大学计算机科学系) Key Laboratory of High Confidence Software Technologies, Ministry of Education(教育部高可信软件技术重点实验室) Big Data Technology Research Center, Nanhu Laboratory(纳米实验室大数据技术研究中心) National Engineering Research Center For Software Engineering, Peking University(北京大学软件工程国家工程研究中心) Peking University Information Technology Institute (Tianjin Binhai)(北京大学信息技术研究院(天津滨海)) School of Computer Sciences, Beijing University of Posts and Telecommunications(北京邮电大学计算机科学系) Huawei Technologies Co., Ltd(华为技术有限公司) Seed, ByteDance Inc.(字节跳动公司)

AI总结 提出LearNAT框架,通过AST引导的分解合成过程和边际感知强化学习,增强小规模LLM的NL2SQL任务分解能力,以7B参数模型达到GPT-4可比性能。

Comments Accepted by ICLR'26

详情

展开后加载摘要…

URL PDF HTML 收藏