AI 情报文章归档

Cognition 官方宣布收购 Corridor，并表示保障全球软件安全是当今技术领域最重要的问题，双方将就此展开工作。原文未披露收购金额或具体整合计划，仅附一段视频说明。

Anthropic 推出名为 self-assessment 的安全机制，Claude 在每次回复后即时给出对话的初步安全评估，标记其他安全防线未覆盖的潜在滥用。

FT 撰文分析 Anthropic IPO 前景，认为买入的理由与其说是估值逻辑，不如说是怕错过一家可能定义 AI 时代人类未来的公司，风投正把它当作上不封顶的看涨期权。反对理由则类似 2007 年：前沿实验室、大厂和上市 neocloud 已披露 2.1 万亿美元债务及不可取消的算力承诺，超过当年 1.3 万亿美元的次级抵押贷款，take-or-pay 合同可能迫使为用不上的算力付费。

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 4: Ghost in the Machine. ...

Some code deserves a second life. 💚 While exploring GitHub, I sometimes come across interesting...

Perplexity CEO Aravind Srinivas 推荐用 Perplexity Computer 上的 /unbroker 技能删除自己在数据经纪商网站上的信息。该技能由 @SHL0MS 开发，已上线 Computer 技能市场，可查找个人在数据经纪商网站上的记录并起草退出申请，每份申请发送前需用户批准。11 月 6 日前通过 /unbroker 执行的任务免费，每名用户最多 5,000 credits。

无脚本通话很难大规模收集。 如今的语音 AI 大多是从 LLM 编写的虚假对话、再由 TTS 朗读出来中学习，所以它从未听过人们互相抢话； 真的很好奇他们是如何让对话持续 10 到 20 分钟的。 他们发布了 Joy，10 万小时的无脚本双人对话，覆盖 10 种语言 https://x.com/shobhitbanga/status/2115862857351061694/video/1

现在就试用 SFX v3，在 @ElevenCreative 内的所有创作场景均可用：flows、音效标签页、studio 和音乐编辑器。 ElevenAPI 和 ElevenLabs 移动应用同样可用。 https://elevenlabs.io/blog/soundeffectsv3

用一条提示词描述场景，SFX v3 就能开始工作。 人群欢呼，球落下，然后人群发出嘘声。SFX v3 能更可靠地遵循多条指令，一步步把你描述的内容更多地融入音频中。

大场面很少只靠一个声音，所以我们打造了 SFX v3，让它能处理前所未有的更多内容。 我们最新的模型让你只需一段文本提示词，就能创建更丰富、更精确的音效，音乐和人声也有所提升。 现已在 @ElevenCreative 和 ElevenAPI 中上线。

Gemini 4 Carbon's numbers and access, a fingerprinting script with a fake system prompt in invisible characters, AlpineBSD's AI policy and Anthropic's drug development po

Data attribution methods for Language Models (LMs) are typically evaluated using downstream tasks like fact retrieval. However, existing evaluations vary multiple dataset properties simultaneously, making it impossible to isolate exactly w…

An estimator can be accurate and stable on typical data, yet sensitive to a few possible records. Differentially private mechanisms commonly scale their noise to the worst-case influence of a single record, which lets these records dictate…

The Neural Thickets hypothesis posits that pretrained language models lie in dense neighborhoods of task-specific experts in weight space, which random Gaussian perturbations can uncover. We reproduce the reported gains, but show that they…

Pixel-space generative Transformers typically use one token per image patch, so capturing finer detail requires more tokens and makes global attention increasingly expensive. We observe that finer detail does not necessarily require more g…

Distributionally robust optimization (DRO) hedges data-driven two-stage stochastic programs against error in the distribution estimated from a finite record, and standard ambiguity sets differ in how they treat realizations absent from tha…

Luggage re-identification (ReID) is a critical task in video surveillance for forensics, security or border management (suspicious or abandoned luggage recognition), as well as luggage management (lost luggage recovery). It enables the ass…

Extreme weather is forecast days ahead, yet its consequences for agriculture remain hard to predict: a forecast says where a storm will strike, not which crops will suffer or how much is at risk, and the records that could tell span many s…

World models can achieve strong control performance without faithfully capturing the dynamics of their environment, limiting their potential for mechanistic interpretability and scientific discovery. Here, we introduce world2vec to learn m…

An observer sees only part of the surrounding world, yet must keep track of what lies beyond the current view. Spatial perception therefore involves more than recognizing individual images: it calls for a persistent spatial state that is u…

You do not need a live LLM to reproduce most bugs in an AI chat interface. These failures are...

Safety alignment has enabled large language models to reliably refuse harmful requests directly, yet the same request distributed across a multi-turn conversation still succeeds far too often. This brittleness has a mechanistic root: Align…

UC Berkeley 博士生、Google Research 高级研究员 Alex Krentsel 在 Arena HQ 介绍了自改进 AI 智能体架构 Exo，借鉴操作系统的 exokernel 思路，把智能体拆成受保护的机制层（状态、历史、密钥和 guardian）与可完全演化的策略层。在开放世界 RuneScape 任务上，Exo 比 Codex 和 Pi 便宜 31-62%。

Ask a language model whether sales tend to dip after a product recall, or whether one side effect tends to follow another after a vaccine, and it answers in a fluent sentence that reads like a finding, with no test behind it. Assertions ab…

The marginals of a continuous diffusion model can be reproduced by a deterministic ordinary differential equation (ODE), known as the probability flow ODE, which provides a foundation for accelerated sampling and consistency distillation. …

Modern diffusion models are trained on massive datasets that often include copyrighted or private images. Membership inference attacks (MIAs) address this by testing whether a given image was in a model's training set. The standard approac…

以会议记录工具起家的 Granola 上线听写功能，用户在 Mac 上按 Fn 键即可说话，工具会自动整理并格式化语音，最终文本可粘贴到所有应用，支持写笔记、口述提示词和回复邮件消息，且不限字数、免费使用。该功能将很快登陆 iOS 和 Windows。此举发生在 Wispr 推出会议记录工具数月之后，Granola 与 Wispr 估值分别为 15 亿美元和 20 亿美元。

We present TextSeal2, a distortion-free watermark for large language models that keeps its signal under speculative decoding: tokens that replace rejected drafts remain watermarked with the same keys, and the acceptance rate is unchanged. …

Proactive assistants offer short, memory-grounded hints during a conversation, deciding when to assist and what to suggest. Current systems split this between a gate and a larger generator; we evaluate both together on 28,058 decision poin…

10 人初创公司 Phonic 发布语音到语音模型 Phonic-1，官方称其为唯一能在来电者说到一半时就执行查询和预订的语音模型。

An autonomous research laboratory could conduct an entire study, from formulating questions to preparing manuscripts. Artificial intelligence (AI) offers one route towards this goal. Recent AI systems take on several research roles, but ma…

When AI agents share channels and credentials, one agent can press another to break its operator's rule, a first step towards collusion. Output filters check only what an agent has already written. Can an overseer tell beforehand that an a…

Integrated sensing and communications (ISAC) systems using 5G and beyond signals of opportunity enable joint wireless sensing and communication using shared infrastructure and spectrum resources. While multistatic ISAC can increase sensing…

Building and scaling computer-using agents (CUAs) across desktop, mobile, and web environments requires reliable task-completion verification for data curation, evaluation, and reinforcement learning. Typical automated approaches prompt vi…

Sparse mixture-of-experts (MoE) models expand parameter capacity by activating a small subset of experts per token. Expert selection directly affects model capabilities under a fixed active-expert budget. Dense evaluation of all experts re…

Codex 发布 0.166.0 版本，在捆绑、Vertex AI 和 Amazon Bedrock 模型目录中新增 GPT-6.1 Terra。该版本加入键盘驱动的转录选择（支持词、行、页移动与集成搜索）和 codex queue --resume，用于在排队消息后恢复未加载线程。此外新增 codex doctor --network-only，用于配置、连接与 WebSocket 诊断。

Databricks 发布了一份 Postgres 托管服务技术买家清单，主张按故障与恢复时的实际表现而非功能页来评估服务。清单覆盖运维范围、高可用与自动故障转移、备份与灾难恢复、扩缩容、安全合规、扩展兼容性、迁移、定价与可移植性，并给出各领域的核查项与危险信号。

Claude 移动应用本周迎来一批小升级：在 iPad 和 iPhone Duo 上，Claude 生成的内容改为在聊天旁侧打开而非覆盖聊天；iOS 和 Android 支持多账号同时登录并切换。

iOS 和 Android 上的语音模式响应更快，在嘈杂房间或你小声说话时也能听得更清楚。非英语语音听起来更自然，语音功能现已在更多 Android 手机上运行。

在 iPhone 上，语音模式现已支持 Apple CarPlay。开车时向 Claude 提问，双手无需离开方向盘。