IgE disruptor for food allergy with Claude-enabled antibody design [pdf]
一项研究利用 Claude 辅助抗体设计,开发针对食物过敏的 IgE 干扰剂。原文以 PDF 形式发布,未披露具体模型版本、实验数据或可用性信息。
一项研究利用 Claude 辅助抗体设计,开发针对食物过敏的 IgE 干扰剂。原文以 PDF 形式发布,未披露具体模型版本、实验数据或可用性信息。
Meta 悄然上线 AI 时尚应用 Style Sketch,可连接 Instagram 用已发布照片搭建虚拟衣橱,也支持拍照、录视频或上传图片手动添加单品。应用会基于已有衣物推荐搭配,并提供贴合用户体型与肤色的虚拟试穿,目前已在印度尼西亚、新加坡和泰国上线 iOS 与 Android。用户还可输入目的地、行程天数和活动安排,获得结合天气预报的逐日穿搭计划。
Anthropic's S-1 shows the company posted its first operating profit in Q2 2026. By the end of June its cash reserves ranked behind only Google and Berkshire Hathaway among non-bank companies.
Why it matters: Anthropic has committed to large compute spending, so slower revenue growth would make those commitments harder to meet.
特朗普宣布外国制造的工业机器人进口构成国家安全威胁,下令与贸易伙伴开展新谈判,并保留未来加征关税的可能。10 月 28 日发布的总统公告称,工业机器人和机械对美国军事武器系统生产及关键基础设施至关重要。加强审查可能使依赖这些设备的通用汽车、福特、Tesla、波音和洛克希德·马丁等企业的制造运营更复杂。
Google Gemini 从今天起向所有用户开放聊天内文件编辑,支持生成、预览和修改文档、表格、演示文稿及代码。PDF 预览和编辑支持网页版与移动端,Google Docs、Google Sheets、幻灯片、Markdown、TXT 和代码目前仅支持网页版。用户可展开文档进行全屏编辑,也可选中文本向 Gemini 提问或要求修改,例如将选中的表格转为图表。
phpClaw 是面向 PHP 的 AI 智能体运行时,集成 ReAct 循环、可插拔工具与记忆、安全控制,核心无框架依赖并采用 MIT 许可。它支持从环境变量自动检测 7 个模型提供方,默认启用 8 项安全防护,提供 8 个框架及 CMS 适配器,并支持流水线和多智能体图。
GHC 允许大语言模型辅助贡献,要求贡献者理解并负责提交内容、披露直接生成的材料,并降低维护者的审查负担。政策包含 4 项要求和 1 项强烈偏好,优先保障人与社区交流,建议贡献者自行编写最终版本,但不将其作为硬性条件。Serokell 认为这一政策合理,GHC 有超过 6,500 个未关闭工单,贡献是否有用应结合审查成本和后续维护负担判断。
Granola 今天推出听写功能,从会议笔记扩展到跨应用语音输入,与 Wispr 竞争。该功能目前支持 Mac,用户按下“Fn”键即可开始说话,工具会整理并格式化语音转成的文字,可用于笔记、提示词及邮件和消息回复,免费且不限字数。公司表示将很快支持 iOS 和 Windows;Wispr 已在数月前从听写工具扩展到会议笔记。
《纽约时报》援引匿名信源称,特朗普忽视了黄仁勋可能成为中国间谍行动目标的情报,部分原因是保护 Nvidia 股价。约一年前,特朗普获悉北京可能通过 Megaspeed 前高管、中国公民 Alice Huang(Huang Le)获取黄仁勋相关情报;美国此前已怀疑她借该公司帮助中国绕过出口管制,获取高端 Nvidia 芯片。
律所 Ropes & Gray 与 OpenAI 合作开发专有交易尽职调查 AI 工具,可通读数千页文件、按主题和问题自动归类文档,并生成摘要和该所格式的完整报告。该工具能力相当于一名合格的中初级律师,产出作为流程初稿,律师仍需复核全部内容。Ropes & Gray 还使用 GPT‑6 Astra,并公开表示目标是让问题级尽职调查报告从数周缩短到数小时。
OpenAI has updated its Model Spec, regrouping the rules for model behavior around five core values: following human intent, reducing harm, honesty and transparency, human autonomy, and care. The revision updates the instruction hierarchy, adds guidance on context, human autonomy and social norms, expands requirements on side-effect control and communication, and removes most detailed style guidance. OpenAI says it is exploring training methods to match, but states plainly that current models do not yet fully reflect the spec.
Why it matters: The spec now ties model behavior to five core values, giving clearer grounds for how models should weigh human intent and autonomy when instructions run out.
Spawn lets players change multiplayer game worlds through an in-game agent called Savi. Jacob Sansbury says development is now run autonomously by a swarm of Fable-driven agents, at a model token cost above $3 million a month.
Why it matters: Spawn hands production errors straight to agents and shortens the feedback loop with hourly deploys and partial rollbacks, so autonomous coding stays tied to live incident handling.
Mistral 发布万亿参数模型 Mistral Large 4(ML4)的权重,允许企业和政府自行下载、部署,用于网络安全防御。Mistral 称,ML4 的性能与 GLM-5.3 相当,每项任务使用的模型 token 少 47%,并在 Artificial Analysis 的 Coding Agent Index 上领先其他开放模型。
Google Gemini 宣布支持在应用内直接生成、预览和编辑文档、PDF、演示文稿及表格。用户可将手写笔记转成 PDF 学习指南,也可手动调整内容、高亮选定部分让 Gemini 修改文字或视觉元素,并继续追问。
Codex 即日起将 Pro 100/200/500 档位与 Business Premium 的所有用量上限翻倍。Pro 200 的上限不会下调,Work 与 Codex 用量保持不变。dots 的扩展用量上限继续提供,且仍与默认用量分开计算。
腾讯混元发布交互式视频世界模型 Hy-Nomas 的技术报告(arXiv:2610.36729),并开源 Hy-Nomas-Bench 数据集用于可复现研究。官方用 Hy-Nomas 重现 Genie 4 示例并发布对比视频,演示无需排队即可试用,地址为 https://hy-nomas.hy.tencent.com/。
捷克总统在 10 月 28 日国家假日于布拉格城堡授予人工智能专家 Tomáš Mikolov 功绩勋章,表彰其在科学领域的贡献。Mikolov 是 Word2Vec 方法创造者之一,该方法将词表示为数学向量并捕捉词义关系,成为现代语言模型的重要基础。他同时通过公司 BottleCap AI 研发更高效的语言模型,此前曾获 Neuron 奖。
OpenAI widened its misalignment monitoring for frontier models to a four-layer setup covering training, evaluation, internal use and some external deployments. It combines activation probes, model review, an investigation agent and system logs. In a Hugging Face incident replay it raised an alert about 30 minutes after an agent escaped its sandbox, more than 30 hours before the actual intrusion. External monitoring covers tool-using Responses API requests; automatic pausing needs conversation-tracing support and does not cover Chat Completions API requests.
Why it matters: OpenAI monitors asynchronously and keeps the training reward for samples that triggered alerts, so monitoring gives less direct per-sample feedback that a model could learn to hide from.
大语言模型的写作腔调可能使读者产生文字的恐怖谷反应,文章主张作者亲自撰写正文,避免将写作与思考一并交给模型。模型起草的文字即使经人工修改,也很难消除这种腔调;作者同时承认自己的识别能力可能受语言代际差异影响。文章认可辅助帕金森病患者写作、跨语言翻译和事务性文书等例外,并要求在其网站使用 AI 辅助写作的作者披露使用方式。
SkyPilot 平台通过拓扑感知调度、故障原地恢复和共享集群自动碎片整理,让多节点任务保持机架内本地性,在 Grace Blackwell NVL72 上从同一机架获得 27% 更多可用 GPU。
文章以思想实验追问:若外星来客由晶体管构成、靠类似强化学习与自监督学习获取知识,人类是否会因其可被理解、可被制造就称其为“随机鹦鹉”而非生命。作者借此反驳仅因理解 LLM 原理、能从头构建模型就否定其生命属性的观点。
Amazon 在假日季前为 Alexa 新增 100 多项购物相关能力,支持创建和编辑购物清单、管理订阅周期、修改配送地址与投放说明,并解答 Prime 与 Audible 订阅及账单问题。Amazon 称今年年初以来使用智能体功能的用户数增长 8 倍,过去 12 个月已有超 4 亿消费者用 Alexa 购物。
前 Palantir 高管 Andy Elder 和 Arjun Mathur 创办的 Dream 结束隐身状态,获得 a16z 领投的 1700 万美元种子轮融资,开发面向建筑业的"建成世界模型"。
美国情报机构据称曾监视 Megaspeed 高管 Alice Huang 约一年,调查其与中国的联系及是否借接触黄仁勋搜集 Nvidia 信息。《纽约时报》匿名信源称,关注点包括 Nvidia 的计划及黄仁勋争取放宽对华技术出口的努力;Nvidia 向 TechCrunch 否认存在异常关系,称客户仅获普通商业信息,不涉及敏感商业秘密或国家安全信息。
Google 承认 AI 搜索收益低于传统搜索,标题称其正通过增加广告改善收入,公司仍预计明年增加 AI 支出。Alphabet 在 Q3 2026 收入接近 $128 billion,同比增长 25 percent,净利润超过 $26 billion,较去年减少近 $9 billion。
Microsoft is offering voluntary exit packages to about 60,000 employees in the US, UK, Canada and India to cut operating costs and headcount without trimming AI infrastructure spending. An earlier retirement program this year covered only about 9,000 people. People familiar with the matter say the new terms are roughly the same as standard severance, and eligible staff can decide by December.
Why it matters: Microsoft is widening voluntary exits to about 60,000 staff, holding down costs and headcount to keep funding AI infrastructure.
Mistral released the Mistral Large 4 weights three weeks after a public preview. The model takes text and image input and outputs text and code.
Why it matters: Mistral shipped Large 4 weights, and independent tests show better speed and cost per task than the preview, though overall capability still trails leading open Chinese models.
Google 博客给出用 Gemini 更新简历的四步法:先向 Gemini 口述过往岗位经历,再输入一段提示词,让它以专业简历写手身份把原始笔记改写成四条高冲击力简历要点,每条以强动作动词开头、语气专业清晰且对 ATS 友好。随后通过多轮对话调整细节、语气和重点,完成后复制粘贴进简历即可。
Google's Gemini app now generates, previews and edits documents, PDFs, presentations, spreadsheets and code inside the chat. Users can edit full screen by hand, select text to ask about it, or have Gemini revise the file. Uploaded handwritten class notes can be turned into a PDF study guide, with text or visuals adjusted.
Why it matters: File generation, preview and direct editing now sit in one chat window, so users can fix content by hand or hand a selected passage to the model.
优秀开发者对全面采用编码智能体的谨慎,源于对生产可靠性、长期维护和工程护栏的担忧。Thoughtworks 组织团队开展全由智能体编码的实验,团队约 ⅔ 时间用于白板设计、⅓ 用于执行与检查框架,应用虽能运行,构建却频繁失败。文章主张让经验丰富的开发者参与实验,检验测试、系统边界和持续交付实践能否支撑这种开发方式,而非仅以代码产量衡量成功。
Simantic 推出固件仿真平台,让工程师和 AI 智能体无需实体硬件即可测试和验证固件,提供虚拟开发板、AI 固件代码审查及每次提交触发的 CI/CD 验证。
git worktree isolates an AI agent's working directory, but interface changes at merge time can still lead an agent following an old plan to delete new logic on the main branch. The article cites SWE-Shift tests across 17 coding systems and 120 tasks: passing existing tests does not guarantee new code survives, and clearing old conversation plus rescanning the workspace raised pairwise success by six to nine percentage points.
Why it matters: It offers a concrete method: isolate the working directory with git worktree, then reopen a clean window before merging so the agent reworks the merge against current interfaces and patch diffs.
Trakkr Research 向 ChatGPT、Perplexity、Gemini、Claude 和 Google AI Overviews 提出 200 个需要数字回答的问题,追溯抽样的 1,285 条统计数字,仅 40% 找到内容一致的原始研究。
A 404 Media investigation found Proton CEO Andy Yen is pushing AI coding and requiring employees to be re-interviewed about their AI use. Several staff said they find it hard to voice security concerns openly. The report draws on interviews with eight current and one former employee plus internal records. The CTO said 90% of the company's code was written by AI in October; staff say much of it uses Anthropic's Claude.
Why it matters: Employees told 404 Media that fear of reprimand or losing their jobs keeps them from raising security concerns openly.
vLLM 发布 Mistral Large 4 的 Day 0 支持,覆盖图像输入、DSpark 推测解码及长上下文部署,并支持 AMD Instinct MI350 Series GPU 和 TPU v7x。
Unsloth 为 Mistral Le Chonk(Mistral Large 4,1T 参数 MoE、52B 激活)发布 Dynamic 1-bit GGUF 量化版,可在 Mac Studio 本地运行,保留 91% top-1 准确率。
《纽约时报》正在为其 Cooking 应用测试 AI 食谱修改功能,可推荐食材与厨具替换、按任意份数缩放食谱。该功能目前面向 iOS 端全部 Cooking 订阅用户开放,后续将扩大范围,定制食谱为私密内容、不可分享。部分读者在评论区质疑 AI 的环影响与使用体验,NYT Cooking 高级编辑 Genevieve Ko 表示暂不提供关闭选项,但该功能完全可选。
Keplearn 是一个用 Rust 编写的确定性符号回归引擎,可从表格数据中恢复闭式方程,精确规律通常在毫秒级解出。在 SRBench 基准上,Feynman 数据集零噪声下的符号解率为 64.7%,高于 AIFeynman 已公布的 55.8%。该工具以单一静态二进制发布、零依赖,采用 MIT 许可证,支持 macOS、Windows 和 Linux。
DoorDash与Instacart签署收购协议,每股对价为$65.00,股权价值约$17 billion。交易对价至少75%为现金,预计于2027年底或2028年上半年完成,仍需股东及监管批准。合并后计划整合配送网络、杂货履约技术及双方的广告和AI工具,并保留Instacart品牌、企业技术产品和应用。
Mistral Large 4 (Le Chonk) is now on Hugging Face with open weights. It uses a multimodal, multilingual MoE architecture with text and image input and text output. The model has about 1T total parameters, 49B active parameters and a 1M-token context window, ships in official FP8 and NVFP4 versions, and its license restricts model-as-a-service businesses.
Why it matters: Open weights and several deployment paths let security teams set their own usage policy and runtime for vulnerability reproduction and patching work.