1
服役超 25 年：国际空间站关键机械臂突发故障，计划由宇航员出舱更换
IT之家 6 月 13 日消息，当地时间 6 月 10 日，NASA 发文称，国际空间站关键机械臂“加拿大臂 2 号”（Canadarm2）在例行操作期间出现故障，目前相关任务已暂停。 NASA 计划于当地时间 6 月 30 日让宇航员在太空行走期间进行在轨更换，好在所需备用件此前已运送至国际空间站…
2
宇瞻推出 GraTherX 内存条散热技术：每侧加厚 0.17mm，故障率降低 60%
IT之家 6 月 13 日消息，宇瞻 (Apacer) 近日宣布推出针对低气流设备设计的内存条散热解决方案 GraTherX，宣称其仅需在裸条模组两侧各加厚 0.17mm 即可 让内存故障率降低 60% 。 GraTherX 面向边缘 AI、工业电脑、嵌入式系统， 采用石墨烯-铜复合材料和一体式双面…
3
A Troubleshooting Log: Proxy Routing, QUIC, Docker State, and Ansible
一次真实的生产故障排查记录，从代理路由到Docker状态，全是踩坑后的硬核经验。
This is not a "how to set up a proxy" tutorial. I run a small self-hosted stack: a Linux client, a proxy client, a VPS, a Dockerized control panel, Xr…
4
GitHub availability report: May 2026
GitHub发布2026年5月可用性报告，详述9起影响服务性能的故障事件，官方透明化运维数据值得关注。
In May, we experienced nine incidents that resulted in degraded performance across GitHub services. The post GitHub availability report: May 2026 appe…
5
DeepSeek API 缓存计费系统出故障后主动向用户退款，赠金已到账
DeepSeek因缓存计费系统数据错误主动向API用户退款并发放赠金，透明负责获认可
IT之家 6 月 11 日消息，今日有开发者反馈称，DeepSeek 开放平台已就其缓存计费系统的数据错误向 API 用户自动返还了相应金额。 根据用户收到的通知，2026 年 5 月 13 日至 2026 年 6 月 9 日期间，DeepSeek 的缓存计费系统出现了部分错误，部分用户的 API …
6
Hoxton 报告称苹果同龄 Intel Mac 故障率是 M 系列芯片 Mac 两倍
IT之家 6 月 11 日消息，英国翻新零售商 Hoxton Macs 于 6 月 1 日发布报告， 称搭载 Apple Silicon 的 M 系列芯片 Mac 的硬件故障率仅为 0.9%，而同龄 Intel Mac 的故障率约为其两倍。 该零售商基于 2013 年以来的维修数据，指出在相同使用年…
7
ProbeLLM: Automating Principled Diagnosis of LLM Failures
自动化诊断大模型失败原因的新框架，用原理性方法定位LLM错误根源。
arXiv:2602.12966v2 Announce Type: replace Abstract: Understanding how and why large language models (LLMs) fail is becoming a central challenge as mod…
8
Scary ChatGPT Bug: AI Generates Nightmarish Images from a Simple Prompt Trick
只需一个简单提示，ChatGPT就能生成让人毛骨悚然的噩梦图像，比恐怖片还吓人！
A newly discovered glitch in ChatGPT is sending shivers down users' spines. By using a simple prompt to retrieve a non-existent image, the AI falls in…
9
Causal Agent Replay: Counterfactual Attribution for LLM-Agent Failures
用反事实归因精准定位LLM Agent失败的根本决策步骤，超越简单启发式方法，让调试不再只看表面。
arXiv:2606.08275v1 Announce Type: new Abstract: When an LLM agent fails -- issues a refund it should not have, calls the wrong tool, leaks data -- exi…
10
Show HN: ARouter – drop-in OpenAI/Anthropic proxy that cuts cost and fails over
一个用Rust写的LLM代理，自动降级到低成本模型可省94%费用，且实现零故障切换。
Article URL: https://github.com/sricola/arouter Comments URL: https://news.ycombinator.com/item?id=48440512 Points: 1 # Comments: 1
11
Show HN: Busbar – every LLM behind one URL, in a single Rust binary
用一个Rust二进制统一多个LLM接口，支持智能池化、故障转移，让调用不同大模型像切换URL一样简单。
I have been working on multiple projects lately involving AI endpoints (including some I run locally) and I found I needed a way to easily load balanc…
12
Ask HN: Is Claude Code down, or its just my region?
Claude Code近期响应缓慢甚至不可用，用户反馈区域问题引发关注
For the last couple of days, Claude Code cli is behaving very slow in responding. Especially after 4.8 got released, but today, it has come down to to…
13
BioBlue: Systematic runaway-optimiser-like LLM failure modes on biologically and economically aligned AI safety benchmarks for LLMs with simplified observation format
系统性揭示LLM在生物与经济安全基准上的类失控优化器失败模式，视角新颖，观察格式简化。
arXiv:2509.02655v3 Announce Type: replace-cross Abstract: Many AI alignment discussions of "runaway optimisation" focus on RL agents: unbounded utilit…
14
POIROT: Interrogating Agents for Failure Detection in Multi-Agent Systems
提出POIROT方法，精准揪出多智能体系统中的故障与幻觉，为安全关键领域AI部署扫清监管障碍。
arXiv:2606.02282v1 Announce Type: new Abstract: Orchestrating Large Language Models into Multi-Agent Systems (LLM-MAS) has unlocked remarkable reasoni…
15
AI in SRE: Where and how Google is deploying agentic AI to improve operations
Google SRE实战揭秘：用agentic AI自动诊断故障、优化运维，效率提升秘籍！
Article URL: https://cloud.google.com/blog/products/devops-sre/how-google-sre-is-using-agentic-ai-to-improve-operations Comments URL: https://news.yco…
16
Early Diagnosis of Wasted Computation in Multi-Agent LLM Systems via Failure-Aware Observability
多智能体LLM系统计算浪费的早期诊断方案，故障感知可观测性提升效率
arXiv:2606.01365v1 Announce Type: new Abstract: Tool-using multi-agent large language model (LLM) systems spend computation through model tokens, tool…
17
FALAT: Tracing Failures in LLM Agent Trajectories via Dependency-Guided Search
LLM Agent轨迹失败定位新方法，依赖引导搜索精准追踪异常路径
arXiv:2606.00765v1 Announce Type: new Abstract: LLM-based agents increasingly solve complex tasks through long trajectories involving reasoning steps,…
18
PropLLM: Propagation-Aware Scene Reconstruction for Network Fault Diagnosis
融合传播感知与LLM的场景重建方法，专为网络故障诊断设计，有望提升根因定位效率。
arXiv:2606.00582v1 Announce Type: new Abstract: Network faults propagate layer by layer along topology and protocol dependencies, yet operations syste…
19
The 5 Systematic Failure Modes of AI Research Reports (and How to Catch Them)
揭示AI研究报告的5种系统性失败模式，帮你揪出单位换算等隐藏错误，提升报告可信度
AI research reports look authoritative. The numbers line up, the charts are clean, and every claim has a source citation. But when you actually open t…
20
广州南站：因贵广客专设备故障影响临时停运列车免费退改签
IT之家 5 月 30 日消息，据 @广州南车站 消息，受贵广客专钟山西至恭城间供电设备故障影响，今天由广州南站开往大理的 G2976 次、开往巫山的 G2250 次、开往桂林北的 G3746 次旅客列车临时停运。铁路部门提示，已购买今天相关停运列车车票的旅客，可通过铁路 12306、自助售取票机、…