论文信息 标题: Adversarial Robustness of Vision in Open Foundation Models 作者: Jonathon Fox, William J Buchanan, Pavlos Papadopoulos 发布日期: 2025-12-19 arXiv ID: 2512.17902v1 PDF链接: 下载PDF 视觉对抗攻击的“矛”...

论文信息 标题: Re-Depth Anything: Test-Time Depth Refinement via Self-Supervised Re-lighting 作者: Ananta R. Bhattarai, Helge Rhodin 发布日期: 2025-12-19 arXiv ID: 2512.17908v1 PDF链接: 下载PDF 照亮深度：Re-Dep...

论文信息 标题: Exploration v.s. Exploitation: Rethinking RLVR through Clipping, Entropy, and Spurious Reward 作者: Peter Chen, Xiaopeng Li, Ziniu Li, et al. 发布日期: 2025-12-18 arXiv ID: 2512.16912v1 PDF...

论文信息 标题: Posterior Behavioral Cloning: Pretraining BC Policies for Efficient RL Finetuning 作者: Andrew Wagenmaker, Perry Dong, Raymond Tsao, et al. 发布日期: 2025-12-18 arXiv ID: 2512.16911v1 PDF链接...

论文信息 标题: Generative Adversarial Reasoner: Enhancing LLM Reasoning with Adversarial Reinforcement Learning 作者: Qihao Liu, Luoxin Ye, Wufei Ma, et al. 发布日期: 2025-12-18 arXiv ID: 2512.16917v1 PDF...

论文信息 标题: Differences That Matter: Auditing Models for Capability Gap Discovery and Rectification 作者: Qihao Liu, Chengzhi Mao, Yaojie Liu, et al. 发布日期: 2025-12-18 arXiv ID: 2512.16921v1 PDF链接: ...

论文信息 标题: DVGT: Driving Visual Geometry Transformer 作者: Sicheng Zuo, Zixun Xie, Wenzhao Zheng, et al. 发布日期: 2025-12-18 arXiv ID: 2512.16919v1 PDF链接: 下载PDF 从无序图像到精确三维：DVGT如何革新自动驾驶的几何感知 论文背景与...

论文信息 标题: EasyV2V: A High-quality Instruction-based Video Editing Framework 作者: Jinjie Mai, Chaoyang Wang, Guocheng Gordon Qian, et al. 发布日期: 2025-12-18 arXiv ID: 2512.16920v1 PDF链接: 下载PDF E...

论文信息 标题: Large Isolated Stripes on Short 18-leg $t$-$J$ Cylinders 作者: Tizian Blatz, Sebastian Paeckel, Ulrich Schollwöck, et al. 发布日期: 2025-12-17 arXiv ID: 2512.15714v1 PDF链接: 下载PDF 量子模拟新突破...

论文信息 标题: Spatia: Video Generation with Updatable Spatial Memory 作者: Jinjing Zhao, Fangyun Wei, Zhening Liu, et al. 发布日期: 2025-12-17 arXiv ID: 2512.15716v1 PDF链接: 下载PDF 从记忆到时空：Spatia如何用3D点云记...