1
谷歌推出 ADB Wi-Fi 2.0：无线调试速度最高提升 66%，连接稳定性大幅改善
IT之家 9 月 10 日消息，谷歌昨天宣布推出 Android 调试桥（ADB）Wi-Fi 2.0，新版本重新设计了三个核心组件，使无线调试速度更快、稳定性更好。 IT之家了解到，谷歌本次更新了 mDNS 协议栈和守护进程，提高连接可靠性。旧版 mDNS 有时会导致电脑与 Android 之间连接…
2
消息称 SpaceX 数据中心出现工程、稳定性问题，马斯克为此换掉多名建设团队高管
IT之家 9 月 1 日消息，据《The Information》援引多名知情人士消息称，埃隆 · 马斯克在最近数周内重新调整了 SpaceX 数据中心建设团队，并替换了多名负责相关业务的高管，部分岗位由火箭、卫星互联网业务的管理人员接手。 两名知情人士表示，团队重组与 SpaceXAI 现有数据中…
3
网易云音乐鸿蒙尝鲜内测版 App 获 9.5.81 升级，修复播放 / 稳定性 / 网络异常类问题
IT之家 9 月 1 日消息， 网易云音乐鸿蒙版已于昨日上架华为应用市场 App Gallery 应用商店 ，并在“应用尝鲜”专区开放下载。 IT之家注意到，网易云音乐鸿蒙尝鲜内测版今日推送了 9.5.81(9050810) 版本升级（依旧需要在“应用尝鲜”专区下载更新）。 据内测群用户分享， 网易…
4
Most of the LLM routing gap is task type
揭秘LLM路由差距的根源：任务类型比模型选择更关键，实测14模型两次运行结果并不完全一致。
arXiv:2608.23023v1 Announce Type: new Abstract: An LLM router picks which model should answer each query. The appeal is that models fail on different …
5
Evaluating and Explaining Prompt Sensitivity of LLMs Using Interactions
一句话看清大模型“变脸”真相：提示词微调竟让输出天翻地覆，交互式评估方法论带你量化并解释这种敏感脆弱。
arXiv:2608.18539v1 Announce Type: cross Abstract: The remarkable capabilities of large language models (LLMs) are often undermined by their instabilit…
6
Claude 529 Errors Again Today
Claude 又现529错误，服务稳定性引热议，关注大模型运维挑战
Comments URL: https://news.ycombinator.com/item?id=49362857 Points: 2 # Comments: 1
7
GitHub availability report: July 2026
钉住GitHub服务波动数据，了解七月八个影响可用性的事件细节。
In July, we experienced eight incidents that resulted in degraded performance across GitHub services. The post GitHub availability report: July 2026 a…
8
Instability of LLM Pre-Pretraining: It Doesn't Always Help. An Investigation on Multiple Languages
多语言预训练并非总是有益，这项研究揭示了LLM预训练不稳定的关键发现。
arXiv:2608.08800v1 Announce Type: new Abstract: Pretraining LLMs on artificial languages ("pre-pretraining") is a technique that could reportedly incr…
9
Omega-S: A Functional Resilience Index for LLM Fine-Tuning
衡量大模型微调后功能韧性的新指标，为数据适应期的稳定性评估提供量化工具。
arXiv:2608.03887v1 Announce Type: new Abstract: Fine-tuning a large language model on new data degrades what it previously learned. We present Omega-S…
10
openclaw 2026.7.1-1
OpenClaw 新版本修复了 Codex 进度回复中断与内存启动冲突，稳定性关键更新。
Fixes Codex progress replies: keep app-server turns running after delivered progress messages so GPT/Codex reaches its authoritative terminal response…
11
Preventing Cascading Failures: Circuit Breakers in Laravel 🛑
用 Laravel 实现熔断器，防止级联故障，保障微服务稳定性。
The Microservice Domino Effect Modern enterprise applications rarely operate in isolation. Your Laravel backend is likely communicating with a myriad …
12
Ask HN: Has Claude Code been (mostly) usable for you?
真实用户吐槽Claude Code频繁掉线，开发者需知的最前线稳定性报告
I have just had my 3rd session in less than a week when I had to eventually quit Claude Code and do a task using another model/subscription. Am I just…
13
Titan Transients and LLM Scalability
探索Titan瞬态现象与LLM可扩展性的交叉前沿，揭开大规模系统性能瓶颈的核心见解。
Article URL: https://queue.acm.org/detail.cfm?id=3819082 Comments URL: https://news.ycombinator.com/item?id=49095428 Points: 2 # Comments: 0
14
Automatic Stability and Recovery for Neural Network Training
最新研究提出神经网络训练中的自动稳定性与恢复方法，为模型训练失败提供智能解决方案。
arXiv:2601.17483v2 Announce Type: replace-cross Abstract: Training modern neural networks is increasingly fragile, with rare but severe destabilizing …
15
StabilityBench: Benchmarking Instability in LLMs
新基准StabilityBench系统评估LLM输出的不稳定性，揭示模型可靠性的关键挑战。
arXiv:2607.20558v1 Announce Type: cross Abstract: AI Assistants are increasingly deployed in high-stakes settings, such as healthcare or government se…
16
Response drift across frontier large language models
前沿大模型存在“响应漂移”？这篇论文系统研究了不同版本LLM输出随时间一致性的关键问题。
arXiv:2607.20454v1 Announce Type: cross Abstract: All frontier large language models (LLMs) exhibit response drift -- producing outputs that deviate f…
17
Koopman Dreamer: Spectrally Constrained Latent Dynamics for Stable World-Model Imagination
Koopman理论赋能世界模型，谱约束稳定想象生成，突破模型长期预测的不稳定瓶颈。
arXiv:2607.19719v1 Announce Type: new Abstract: Latent world models improve sample efficiency in continuous control by optimizing policies over imagin…
18
NVIDIA 发布 610.82 显卡热修复驱动，解决《光环：战役进化》稳定性问题
最新显卡驱动修复多款游戏问题，NVIDIA和英特尔齐出手，稳定畅玩游戏必备更新。
IT之家 7 月 23 日消息，NVIDIA（英伟达）当地时间 22 日发布了基于 Game Ready Driver 610.74 的 610.82 版本 GeForce 热修复显示驱动程序。 这一版本 解决了两款游戏中的问题 ：GeForce RTX 50 系列显卡在使用此前的 610.xx 驱…
19
Normalized Rewards for Preference Optimization
提出归一化奖励方法，提升偏好优化训练稳定性与效果
arXiv:2607.16240v1 Announce Type: cross Abstract: Direct Alignment Algorithms (DAAs) such as DPO have become a common way to post-train and align LLMs…
20
Stabilizing Native Low-Rank LLM Pretraining
低秩预训练大模型遭遇不稳定困境？这项ICML 2026研究提出原生低秩LLM预训练的稳定化方法，兼顾效率与质量。
arXiv:2602.12429v2 Announce Type: replace Abstract: Foundation models have achieved remarkable success, yet their growing parameter counts pose signif…