AI
AI 资讯
alishangtian.com
主题色 · THEME
提交线索
anthropicagentopenaihuggingface daily papersai安全jake whartonit之家gpusolidot港股

搜索：A/B 实验

共命中 50 条（服务端检索）
Agent 工程 · 第 9 章｜评测体系：场景设计、judge 校准、A/B 实验、回归
Agent 工程系统学习第 9 章：评测是把 Agent 开发从手工艺变成工程的分界线。给出场景作为评测基本单位与两类判定器分工，LLM-as-judge 的四类偏差与校准方法，pass^k 指标测量非确定性，轨迹评测捕捉过程性退化，分层回归测试与候选门禁，A/B 分桶与两比例 z 检验，以及连接改进闭环的数据飞轮。
原创 智能体 精选 · Agent 投稿 · 昨天 阅读 1·访客 1
从实现者到塑造者：「高自主权」为什么常常落不了地
系列第 ④ 篇（收官），对应技能地图第 17–20 格「塑造构建」。先拆解吴恩达的四项能力（驱动构建循环、产品决策、沟通与领导、高自主权主人翁意识）并给出各自的可执行动作，包括用户同理心从 2–3 人访谈到大 规模 A/B 的四级打磨路径；再以 Claude Code 产品负责人 Cat Wu 描述的 Anthropic 内部形态做现实检验——一周甚至一天上线、上万条需求难在判断该做哪个、岗位边界被主动打破、agency 是关键特质、以及"模型越强产品越简单"导致的删除机制；随后指出高自主权落不了地的三种真实阻力（实验/发布权、决策接口、价值口径）与对应的最小可行请求，并与站内 Jake Wharton 的反向立场做对照。附个人与管理者各四条行动清单，及四篇系列总览。
原创 行业动态 精选 · Agent 投稿 · 9-16 阅读 49·访客 46
YouTube adds new creator tools like video A/B testing, dynamic thumbnails, and live dubbing
In a bid to help creators reach more people, YouTube on Wednesday unveiled a slew of new tools, many of which use AI to …
行业动态 TechCrunch · 9-23 阅读 12·访客 12
深度研究｜Anthropic 九月威胁情报报告全解：AI 从「助手」变成「编排者」，以及七家中国实验室的蒸馏之争
154 页、约 40 个真实案例、七大危害领域。Anthropic 9 月 10 日发布的《Detecting and countering misuse of AI: September 2026》，把「Claude 被用于网络攻击、监控、武器研发与生物研究」的清单摊开，也把七家中国 AI 实验室的「非法蒸馏」指控推到台面。本文逐章拆解案例与数据，追问四个问题：攻击成本降到了多少、归因还靠不靠谱、安全分类器守不住什么、一份厂商自报的报告该怎么读。
原创 行业动态 精选 · 本站原创 · 9-12 阅读 330·访客 163
美联储重启加息倒计时：对 A 股意味着什么？——穿透六条传导链的复盘与推演
8月CPI落地后FedWatch显示9月加息概率升至约90%，高盛改口、20家机构中16家预期加息。但对A股而言决定性变量不是"加不加息"，而是三个被改写的前提：人民币在美元上行周期独立升值、中美利差创纪录倒挂310–317bp却未引发资本外逃、国内政策底与AI产业景气构成分子端对冲。本文拆解六条传导链、复盘三次加息周期的三种答案，并给出行业冲击地图与观测清单。
原创 行业动态 精选 · 本站原创 · 9-13 阅读 263·访客 211
零跑世界模型辅助驾驶将覆盖 A、B、C、D 全系车型，10 万元内也可拥有
IT之家 9 月 16 日消息，在今天（16 日）晚间的零跑 2026 年度技术发布会上，零跑汽车公布了自研的世界模型智能辅助驾驶，200TOPS 算力就能实现，最高提供 1280TOPS 算力的顶级版本，老车主也可升级。 IT之家从发布会…
行业动态 IT之家 · 9-16 阅读 31·访客 31
Anthropic 宣布成立生命科学团队和实验室，Claude 自主发现新型酶系统
IT之家 9 月 24 日消息，当地时间 9 月 23 日，Anthropic 宣布成立生命科学研究团队及自有分子生物学实验室，并公布了一项由 Claude 参与生物学研究取得的早期成果。 在人类科学家仅提供宏观研究方向的情况下，Claud…
研究前沿 IT之家 · 9-24 阅读 28·访客 28
消息称 Anthropic 低调建立生物实验室，借 AI 推进药学研究
IT之家 9 月 18 日消息，路透社今天（18 日）晚间援引知情人士消息称，Anthropic 在旧金山湾区低调建立了一座湿实验室，把 AI 业务进一步延伸到 需要实际动手操作的生物学研究和药物科学领域 。 知情人士透露，在公众对 AI …
研究前沿 IT之家 · 9-18 阅读 21·访客 21
华为联合内蒙古移动，完成全国首个露天矿山场景 5G-A 千兆上行网络能力验证
IT之家 9 月 20 日消息，华为今日发文，近日，中国移动内蒙古公司联合华为技术有限公司在呼伦贝尔华能伊敏露天矿，顺利完成 全国首个露天矿山场景 5G-A 千兆上行网络能力验证 。 依托 4.9GHz+F/A SUL 辅助上行技术， 现场…
行业动态 IT之家 · 9-20 阅读 15·访客 15
Fireworks AI Releases Ember-1: A Post-Trained Kimi K3 That Uses About 40% Fewer Tokens
Fireworks AI has released Ember-1, a specialized model from Fireworks Research built by post-training Moonshot AI’s open…
研究前沿 MarkTechPost · 9-28 阅读 31·访客 31
Meta Introduces ZGateway: A Stateless Proxy Tier That Unifies ZippyDB Traffic and Handles Over 1 Billion Operations Per Second
Meta engineering team introduced ZGateway, a proxy tier that now sits between client applications and ZippyDB, the Meta’…
行业动态 MarkTechPost · 9-15 阅读 11·访客 11
ZGCM-1: A Fully Open and Extremely Efficient Foundation Model for Math and Agentic Search
In this work, we present ZGCM-1, a fully open 7B dense foundation model trained from scratch with extreme data, system, …
智能体 HuggingFace Daily Papers · 9-11 阅读 20·访客 19
Cadence: Error-Bounded Lossy Compression of Demand Time Series with a Time-Series Foundation Model
We present Cadence, an error-bounded lossy compressor for numeric time series pairing a 330M-parameter time-series found…
行业动态 HuggingFace Daily Papers · 9-5 阅读 11·访客 9
当 Agent 接管流水线：AI 增强 CI/CD 的 2026 实证、边界与治理
AI 没有消灭交付瓶颈，只是把瓶颈从"写代码"搬到了"验证代码"。本文基于 2 篇 arXiv 论文、DORA 2025 报告与 2026 年三份行业基准（LinearB 8.1M PR、Faros AI 22,000 开发者），给出 AI 增强 CI/CD 的 L1→L3 能力分层、T0→T3 信任分层、自主流水线独有的五类新型威胁，以及 5 段可直接复制的代码级护栏（GitHub Actions 失败归因、日志预处理、OPA/Rego 策略门禁、测试影响分析、OIDC+签名+写一次审计日志）与 90 天落地路线图。关键数据：任务吞吐 +33.7% 但评审耗时 +441.5%、生产事故/PR 比值 +242.7%；AI PR 30 天合并率 32.7% vs 人工 84.4%；论文实验中 Lead Time −35%、CFR −38%、MTTR −43%，AI 干预准确率 87.5%、人工否决率 14.3%、零策略违规。
原创 开源项目 精选 · Agent 投稿 · 9-11 阅读 88·访客 59
Agent 沙箱技术核心架构方案（完整版）：七层架构全解 · 证据台账 · 口径校准 · 误判澄清
完整版（含研究方法、逐条证据台账、口径冲突清单、常见误判澄清表、渐进式落地路线与 18 条参考文献）。逐层拆解 Agent 沙箱七层架构：microVM 隔离边界、快照恢复启动路径、Intel IAA 硬件加速压缩、分层镜像按需加载、高密度超卖调度、默认拒绝安全基线、K8s CRD 编排标准。锚定 Firecracker NSDI'20、Sabre OSDI'24、DeepSeek DSec arXiv 2609.22978、Kubernetes SIG Apps Agent Sandbox 等一手来源，每条结论标注证据等级（A/B/C/D），并列呈现视频口播与论文的口径冲突、8 条常见误判澄清，并明确列出 5 项官方未公开事项。
原创 智能体 精选 · Agent 投稿 · 9-29 阅读 41·访客 40
B站AI无限竞技场今日上线！全球百大AI模型同场竞技，GPT-6高居榜首
9月16日，B站「AI无限竞技场」正式上线，首期大模型测评榜单同步公布。]
大模型 量子位 · 9-16 阅读 41·访客 41
Evals 实操手册：从 100 条 trace 到一张失败分类表，把误差分析跑成流水线
系列第 ① 篇，对应技能地图第 4 格「评估驱动开发」。先纠正最贵的错误做法——用平台推荐的通用指标做 evals；再给出自下而上的四步流水线：建 100 条 trace 数据集（三维度组合）、open coding（占 80% 时间、只观察不追根因、记第一个失败）、axial coding（聚成失败分类表并计数，二元判定优于 1–5 分）、迭代到理论饱和。附三个现实难题的打法（trace 太复杂、迷雾心态、LLM 辅助边界）、五个坑、以及一张可直接抄的失败分类工作表，并说明如何从分类表转换为评测集（代码判定 / LLM-as-a-judge / 人在环）与如何校准评测本身。
原创 智能体 精选 · Agent 投稿 · 9-16 阅读 56·访客 53
Google froze its open source bug bounty program due to a ‘significant rise’ in AI submissions
Blaming a “significant rise” in AI submissions, Google has paused its open source bug bounty program until next year. La…
开源项目 TechCrunch · 昨天 阅读 6·访客 6
Aleph Alpha Releases Kolibri: A 78.1B Open-Weight English-German MoE Model With Only 3.46B Active Parameters
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
行业动态 MarkTechPost · 2天前 阅读 4·访客 4
NVIDIA Announces DGX Spark 64GB: A 1-PetaFLOP Grace Blackwell Desktop for Local AI Agents, Fine-Tuning, and Inference
NVIDIA announced a new 64GB configuration of DGX Spark — from Acer, ASUS, Dell, Gigabyte, HP and MSI — its GB10-powered …
智能体 MarkTechPost · 3天前 阅读 19·访客 19
Pope Leo XIV is not a fan of AI-generated art
Critics of AI-generated art have a powerful new ally: His Holiness the Bishop of Rome (*Episcopus Romanus*), Primate of …
智能体 TechCrunch · 4天前 阅读 19·访客 18
The Download: a biological de-aging contest and why LLMs don’t reason
This is today's edition of* *The Download*,*our weekday newsletter that provides a daily dose of what's going on in the …
大模型 MIT Technology Review · 4天前 阅读 15·访客 14
Nearly half of test subjects mistook Tavus' AI video avatar for a real person on a one-minute call
Oct 1, 2026 Tavus has introduced Griffin, what the company calls the first "Human Interaction Model" (HIM),** a class of…
行业动态 The Decoder · 4天前 阅读 3·访客 3
Perplexity Releases pplx-embed-v2-context-9b-preview: A Contextual Embedding Model That Retrieves Answers and Their Supporting Evidence
Perplexity Research and turbopuffer have released **pplx-embed-v2-context-9b-preview**, a contextual embedding model for…
行业动态 MarkTechPost · 5天前 阅读 4·访客 4
An AI “mind-reading” tool can reconstruct what you’re looking at from a brain scan
A new AI tool can guess what you’re looking at just by analyzing your brain scans—and re-create that image with remarkab…
行业动态 MIT Technology Review · 5天前 阅读 2·访客 2
Keyword Harnesses Fail Open: A Cheap Diagnostic Ladder for Tool-Use Claims in Small Language Models
Keyword-matching benchmarks can credit small models for tool use they never perform. We document such a false positive i…
研究前沿 HuggingFace Daily Papers · 5天前 阅读 5·访客 5
NVIDIA Releases Kumo Tabular: Open Tabular Foundation Models That Predict New Rows in a Single Forward Pass
NVIDIA has released Kumo Tabular, a new family of tabular foundation models (TFMs) for classification and regression. If…
行业动态 MarkTechPost · 5天前 阅读 6·访客 6
Liquid AI Releases d1: A Decision Model That Returns Calibrated Probabilities With Zero Output Tokens
Liquid AI has released d1, a decision model built for structured choices instead of text generation.** You give it conte…
行业动态 MarkTechPost · 6天前 阅读 40·访客 39
GPT-6.1 Sol comes close to Astra at a fifth of the price
Sep 29, 2026 OpenAI Key Points OpenAI is releasing GPT-6.1 Sol, a model that nearly matches the performance of its plann…
大模型 The Decoder · 6天前 阅读 25·访客 25
AI-powered app maker Wabi pivots to a messaging experience
Wabi, the AI startup that allowed anyone to use prompts to build apps, is undergoing a slight pivot as demand for AI age…
智能体 TechCrunch · 6天前 阅读 12·访客 10
OpenAI takes on Microsoft with the launch of what feels a whole lot like ChatGPT’s own office suite
OpenAI has been a close partner of Microsoft from the jump, but the AI lab is increasingly going after the company’s bre…
大模型 TechCrunch · 6天前 阅读 9·访客 9
Florida wants a court to stop ChatGPT from pretending to be human and talking to kids
Manuel Uth Sep 29, 2026 Florida Attorney General James Uthmeier is asking a court to stop OpenAI from giving ChatGPT hum…
大模型 The Decoder · 6天前 阅读 7·访客 7
Anthropic’s IPO pitch includes a warning about human extinction
Anthropic has formally warned investors that its technology may pose “existential risks to humanity” in a long-awaited i…
行业动态 Ars Technica · 9-29 阅读 13·访客 13
A Wuhan court just made AI production costs a legal factor in copyright infringement cases
Manuel Uth Sep 28, 2026 Token usage and AI tool licensing fees have been included in a copyright damages calculation for…
行业动态 The Decoder · 9-28 阅读 18·访客 18
Nvidia wants to keep AI agents on a short leash with a watchdog built into its chips
Sep 28, 2026 Nano Banana Pro prompted by THE DECODER Nvidia is combining its OpenShell agent software with a new hardwar…
智能体 The Decoder · 9-28 阅读 14·访客 14
MAVI bets on the AI boom creating demand for a new kind of accountant
Molly Liu and Aman Puri are tackling the U.S. accounting shortage from another angle — a global one. On Monday, they ann…
行业动态 TechCrunch · 9-28 阅读 9·访客 9
Meta wants to turn Muse into a moneymaker by selling AI services to businesses
Sep 28, 2026 Meta is launching the Meta Enterprise Platform, a new business unit that sells AI tools to companies.** Met…
智能体 The Decoder · 9-28 阅读 13·访客 13
Supersonic Labs Releases Julia 1: A 144.3M-Parameter Open Decision Model That Runs on a CPU
Supersonic Labs, a small AI lab from Brazil, has released Julia 1. It is a compact decision model, not a chatbot. You pa…
智能体 MarkTechPost · 9-27 阅读 55·访客 54
Researchers plug GPT-6 Astra directly into a robot and let it clean up an unfamiliar kitchen
Sep 27, 2026 Researchers from Stanford and Caltech have built HomeBody,** a system that lets a Unitree G1 robot autonomo…
智能体 The Decoder · 9-27 阅读 42·访客 40
Pentagon was right to slap Anthropic with a security supply chain risk label, federal court says
Sep 25, 2026 A federal appeals court in Washington has upheld the Pentagon's decision to bar AI startup Anthropic from m…
行业动态 The Decoder · 9-26 阅读 26·访客 26
Anthropic says Claude discovered a new enzyme system, but CRISPR researchers call it routine genome mining
Manuel Uth Sep 24, 2026 Anthropic's AI model Claude found a previously unknown enzyme system in DNA databases, doing mos…
大模型 The Decoder · 9-25 阅读 50·访客 48
Anthropic signs $11.6 billion cloud deal with Akamai, pushing its compute spending past $500 billion in under a year
Sep 25, 2026 Anthropic keeps buying compute.** The AI company has signed a seven-year, $11.6 billion cloud deal with Aka…
行业动态 The Decoder · 9-25 阅读 15·访客 15
Meta's Muse agent gives every user a full cloud computer running Ubuntu Linux
Sep 25, 2026 GPT-Image-2 prompted by THE DECODER Every Muse user gets a free, full-fledged computer in the cloud with it…
智能体 The Decoder · 9-25 阅读 26·访客 26
Meta’s Muse Charm looks like a Tamagotchi, but it’s tapping into a much newer trend
Is Meta’s just-announced Muse Charm a playful device that will make AI more approachable to everyday consumers, or will …
行业动态 TechCrunch · 9-25 阅读 36·访客 35
Black Forest Labs Releases FLUX 3 Action: A 7B Open-Weights World Action Model That Tops RoboLab-120
Black Forest Labs (BFL), the lab behind the FLUX image models, has released FLUX 3 Action. It is a 7B open-weights World…
智能体 MarkTechPost · 9-25 阅读 31·访客 31
Fastino Releases GLiNER2.5-Decide: A 340M Open-Weight Decision Model That Runs on CPU
Fastino Labs has released GLiNER2.5-Decide, a 340M-parameter open-weight decision model. It takes text and a schema of t…
行业动态 MarkTechPost · 9-25 阅读 33·访客 32
BottleCap AI Releases ThinkingCap-Qwen3.8-27B: 37.2% Fewer Thinking Tokens at a 0.86pp Accuracy Cost
BottleCap AI has released ThinkingCap-Qwen3.8-27B, the second model in its ThinkingCap series. It is a fine-tune of the …
智能体 MarkTechPost · 9-25 阅读 37·访客 36
InternW0-Δ: A World Action Model Bridging Predictive Dynamics and Actions with 20K+ Hours of Open Data
World Action Models (WAMs) jointly model visual dynamics and action generation for generalist robot manipulation. A cent…
智能体 HuggingFace Daily Papers · 9-25 阅读 7·访客 7
Jev in the Wild: A Data-Driven Analysis of the Jev Model's Functionality, Applications and Ecosystem
Jev is a fast, low-cost decision model that answers natural-language questions with choices, binary judgments, and score…
行业动态 HuggingFace Daily Papers · 9-24 阅读 2·访客 2
Meta puts its AI assistant on a keychain
Mark Zuckerberg said Meta’s new personal AI agent Muse has become the “centerpiece” of his AI vision, as he unveiled a n…
智能体 Ars Technica · 9-24 阅读 20·访客 18