1
Anomaly-Preference Image Generation
ICML 2026前沿成果，提出异常偏好图像生成新范式，为可控生成开辟新方向。
arXiv:2605.02439v2 Announce Type: replace-cross Abstract: Synthesizing realistic and diverse anomalous samples from limited data is vital for robust m…
2
Leveraging Unsupervised Learning for Cost-Effective Visual Anomaly Detection
探索无监督学习实现低成本视觉异常检测，新方法兼顾效率与精度。
arXiv:2409.15980v2 Announce Type: replace Abstract: Traditional machine learning-based visual inspection systems require extensive data collection and…
3
Domain Module Postmortem
Node.js 官方深度复盘 Domain 模块的设计缺陷，揭露其隐式异常捕获如何破坏模块隔离性。
4
VLM-AutoDrive: Post-Training Vision-Language Models for Safety-Critical Autonomous Driving Events
这篇论文提出了模块化后训练框架，将视觉语言模型适配到自动驾驶安全关键事件的异常检测，融合元数据、LLM描述与VQA及CoT推理，提升准确性。
arXiv:2603.18178v2 Announce Type: replace Abstract: The rapid growth of ego-centric dashcam footage presents a major challenge for detecting safety-cr…
5
PaAno: Patch-Based Representation Learning for Time-Series Anomaly Detection
提出PaAno方法，用补丁表示学习实现高效时间序列异常检测，兼顾性能与计算资源
arXiv:2602.01359v2 Announce Type: replace-cross Abstract: Although recent studies on time-series anomaly detection have increasingly adopted ever-larg…
6
Tell HN: Audible app used 19.8GB of data while not being used
Audible静默偷跑20GB流量，用户反馈无门，值得警惕的软件异常行为。
Seriously, Audible used nearly 20GB of data in a day, while not being used. I listened to a downloaded book in the morning. Left my phone on my desk. …
7
Unsupervised Domain Shift Detection with Interpretable Subspace Attribution
机器学习新工具：无监督检测域迁移，还能定位异常特征子空间，实现可解释归因。
arXiv:2605.15920v1 Announce Type: cross Abstract: We developed a tool for detecting domain shifts, namely subtle differences in the probability distri…
8
DeepSeek 回应 <think> 字符触发模型异常：系特殊字符引发幻觉，不涉及安全问题或隐私泄露
DeepSeek 官方回应 <think> 字符触发异常：系模型幻觉，非安全问题或隐私泄露。
IT之家 5 月 19 日消息，今日 DeepSeek 就“<think> 字符触发模型异常回复”发布说明，官方称：属于特殊字符引发的模型幻觉，不涉及安全问题或隐私泄露。 IT之家附声明全文如下： 关于 近期，我们关注到有用户反馈，在与 DeepSeek 模型的对话中输入“<thi…