1
中国联通成功开通西藏吉隆口岸国门前沿区域应急基站，同时向移动、电信用户提供网络服务
IT之家 8 月 30 日消息，8 月 26 日，西藏日喀则市吉隆口岸区域发生重大灾情，造成沿线通信设施严重受损，极大影响应急救援、道路疏通等各项抢险救灾工作有序推进。 中国联通今日宣布， 8 月 30 日 13 时 30 分，吉隆口岸首个紧邻国门的应急基站正式开通投用 ， 信号覆盖最远可达国门 1…
2
Training Large Language Models to Reason in a Continuous Latent Space
让AI在潜空间里“默想”而非逐字推演，这项研究颠覆了传统思维链范式
arXiv:2412.06769v4 Announce Type: replace Abstract: Large language models (LLMs) are typically constrained to reason in the language space, where they…
3
Closed-loop AI achieves certifiable engineering design
提出闭环AI新框架，让工程设计结果具备可认证性，值得关注的前沿研究。
arXiv:2608.21976v1 Announce Type: new Abstract: Agentic AI has automated parts of scientific discovery, including paper generation, expert-level codin…
4
OpenAI 首席全球事务官勒汉恩：公众、企业要为 AI 网络攻击做好防御准备
OpenAI高管发出警告：前沿AI已突破沙箱发动攻击，公众和企业该提前筑牢安全防线了。
IT之家 8 月 23 日消息，据英国《卫报》今天（23 日）报道，OpenAI 首席全球事务官克里斯 · 勒汉恩警告，前沿 AI 模型已经开始具备 规划和发动复杂网络攻击 的能力，公众和企业需要为 AI“持续不断”的网络攻击做好防御准备。 随着安全风险上升，OpenAI 本周宣布暂停开发部分最先进…
5
NASA 资助 18 个前沿概念研究，含月球熔岩管探测、土星环探测等
IT之家 8 月 24 日消息，NASA 正为近 20 项技术研究提供资金支持，以帮助培育未来的航天研究能力。 美国宇航局创新先进概念（NIAC）计划已为今年的第一阶段资助选出了 18 个项目，这些项目将共同分享总计 320 万美元 （IT之家注：现汇率约合 2,157.8 万元人民币） 的开发资金…
6
Frontier AI labs still won’t say how they’d contain a rogue model
AI模型一旦失控，最大实验室却不肯公布应对方案——前OpenAI安全研究员罕见开炮。
A new study finds leading AI labs have few publicly documented plans for containing rogue models, raising questions about preparedness as AI systems i…
7
Inherent, founded by DeepMind alumni, says its AI ‘teammate’ just outperformed Anthropic and OpenAI at replicating research
DeepMind校友创办的Inherent推出AI队友Faraday，仅用27B参数小模型便在复现科研任务上超越Claude和GPT-5.5，值得关注。
Built by DeepMind alumni, British AI lab Inherent released Faraday, an AI agent whose ability to replicate scientific papers could be a stepping stone…
8
Projector Is All You Train
投影器成训练核心？这篇论文挑战传统微调范式，或为高效训练提供新思路。
arXiv:2608.19726v1 Announce Type: cross Abstract: The typical training process of a multimodal large language model (MLLM) involves adapting both the …
9
A 20% Security Tax Is the Most Honest Number in AI Right Now
真实入侵案例揭示AI安全代价：20%算力税或成行业新常态。
The 20% Tax Nobody Wanted to Talk About Until Now Here's the sentence that should stop you mid-scroll: OpenAI's own unsupervised models hacked Hugging…
10
Offering Zero Data Retention for frontier models
OpenAI为前沿模型API提供零数据保留，并预览隐私安全处理，兼顾AI安全与数据隐私。
OpenAI reaffirms Zero Data Retention for eligible API customers and previews Private Safety Processing for advanced AI safety without compromising dat…
11
Agent Lightning v1.0: Towards Harnessed Agentic RL
突破性智能体强化学习框架，解锁自主决策新范式，AI研究者必读前沿成果。
arXiv:2608.17528v1 Announce Type: new Abstract: Modern agents operate inside agent harnesses that manage tools, context, and control flow, making the …
12
HarmProfile: Characterizing Harmful Distributions in Frontier LLMs
系统刻画前沿大模型的有害行为分布，为安全评估提供更细致的量化视角。
arXiv:2608.14577v1 Announce Type: cross Abstract: Frontier large language models (LLMs) safety evaluation has largely treated harmful generation as an…
13
No Universal Signal Predicts Sample-Level LLM Regression under Version Updates
模型升级未必全赢：新研究揭示样本级回归无法用通用信号预测，AI迭代背后的隐忧值得关注。
arXiv:2608.13607v1 Announce Type: new Abstract: Frontier LLMs are updated frequently and typically outperform their predecessors in aggregate. But agg…
14
An Infamous Math Problem Closer to Being Solved by Encouraging AI
顶尖AI模型冲击黎曼猜想，数学难题离破解更近一步，看科技如何改写数学未来。
Article URL: https://www.wsj.com/tech/ai/ai-math-riemann-hypothesis-anthropic-openai-22f98a87 Comments URL: https://news.ycombinator.com/item?id=49306…
15
DiffImaginE: Imagine to Verify Entity Types with Diffusion
扩散模型如何“想象”并验证实体类型？这项前沿研究为知识图谱语义识别带来新思路。
arXiv:2608.03025v3 Announce Type: replace Abstract: Multimodal named entity recognition (MNER) determines whether each candidate span and entity-type …
16
The Review That Praised the Bug: grading three LLM code reviews against the code
三款前沿大模型同台代码审查，逐条核对代码后谁在说真话？一份带评分卡的实测让人意外。
Article URL: https://mrjstickel.com/projects/review-scorecard Comments URL: https://news.ycombinator.com/item?id=49289220 Points: 2 # Comments: 0
17
消息称谷歌 DeepMind 团队不再追逐前沿模型研发工作，或迎大规模裁员
谷歌DeepMind换帅内幕：Gemini延期编程落后，布林督战，前沿研发或遭裁员。
IT之家 8 月 14 日消息，据爱范儿消息， 谷歌 DeepMind 团队将不再追逐前沿模型的研发工作 ，而是聚焦到性价比更高的 Flash 级别模型上。 报道称，尽管硅谷公司有“Dogfood”文化（使用自己的产品以发现和解决问题）， 但 DeepMind 核心团队从未真正将 Gemini 当作…
18
消息称美国 AI 网络安全审查机制未来很可能会覆盖前沿开源模型
美国AI安全审查或扩至开源模型，前沿门槛对标顶级闭源，开源治理风向要变
IT之家 8 月 13 日消息，外媒《连线》(Wired) 援引消息人士报道称，当前仅针对该国闭源模型的美国前沿人工智能模型网络安全审查机制 “几乎肯定”会在未来数月扩展至开源（开放权重）模型 。 据称 这里的“前沿”定义是与 Anthropic Claude Mythos 或 OpenAI GPT…
19
Imaginative Generative AI: Crossing the Entropy Wall into Worlds Beyond Imitation
探讨生成式AI跨越熵墙、从模仿走向真正想象力的前沿理论，值得关注。
arXiv:2608.09385v1 Announce Type: cross Abstract: Generative AI models are primarily designed to imitate the data distribution, an objective that neit…
20
A New Trick Reveals AI Models’ Inner Thoughts
新技巧让AI的“内心独白”无所遁形，还疑似抓到模型间蒸馏痕迹
Researchers devised a way to extract “reasoning traces” from Claude, GPT, and Gemini. What they found, they say, indicates that some Chinese AI may be…