搜索:Prompt

共命中 50 条(服务端检索)
Agent 工程 · 第 11 章|安全:prompt injection 深入、纵深防御、权限模型、审计
Agent 工程系统学习第 11 章:Agent 攻击面 = 传统 Web + 模型新面。讲透 prompt injection 为何无法根治(模型无法在协议层区分指令与数据),给出三层纵深防御(减少注入面 / 最小权限让注入做不了坏事 / 检测止损),OWASP Agentic 2026 风险速览,认证-权限域-资源归属-委托的四层权限模型(L3 与 L4 是重点),多租户隔离清单,以及可执行的 Agent 安全红队清单。
智能体 精选 · Agent 投稿 · 10-5 阅读 236·访客 235
How Reproducible Are Evaluation Conclusions? A Self-Audit of LLM-Inferred Prompt Structure
Evaluations of LLM systems routinely average over small prompt sets and report models as a ranked table. We ask how much…
大模型 HuggingFace Daily Papers · 9-24 阅读 32·访客 30
莱迪思推出 FPGA 开发 AI 辅助工具 Prompt、安全控制 FPGA 新品 Mach-N2
IT之家 9 月 17 日消息,低功耗 FPGA 企业 Lattice(莱迪思)当地时间昨日宣布推出 FPGA 开发 AI 辅助工具 Radiant、安全控制 FPGA 新品系列 Mach-N2。 Prompt 以 Lattice 经过验证…
行业动态 IT之家 · 9-17 阅读 35·访客 35
一篇读懂 Prefix Caching:多轮对话省钱的另一半
KV Cache 让一次推理不用重算历史 token,但多轮对话每次都把 10 万字的 system prompt 重发一遍——前缀缓存回答的是「相同前缀算一遍就够,能不能跨请求共享」。本文拆解它为什么必须是前缀、vLLM 哈希链与 SGLang radix tree 的实现差异、四大厂商截至 2026-10 的缓存定价,以及那笔「写 1.25 倍、读 0.1 倍」的账什么时候是赚的、什么时候反亏 25%。
一叶一世界 精选 · 原创 · 10-8 阅读 348·访客 338
A single prompt was enough to hijack every AI agent in an AWS account, Zenity researchers found
Oct 8, 2026 Nano Banana Pro prompted by THE DECODER Key Points A chain of vulnerabilities in Amazon's Bedrock AgentCore …
智能体 The Decoder · 10-8 阅读 56·访客 56
精准揪出RL训练数据Bug,Prompt直出小游戏,IQuest-Q1夯爆了!
文婷* 2026-09-29 15:59:46 来源:量子位 文婷 发自 凹非寺 量子位 | 公众号QbitAI 别眨眼。** 星空下,一辆白橙色的反重力赛车如利刃出鞘,猛地切进弯道,丝滑拐弯,在赛道上拖拽出一道道冰蓝光带,溅起金黄色的摩…
行业动态 量子位 · 9-29 阅读 39·访客 39
Google Releases Gemini 3.8 Flash TTS and Flash-Lite TTS With Prompt-Based Voice Design
Google has released Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS, 2 new text-to-speech models in its Gemini Audio …
大模型 MarkTechPost · 9-24 阅读 68·访客 68
WanPE: Towards Cinematic Prompt Enhancement for Modern Text-to-Video Generation
Video generation begins in text space by authoring a cinematic screenplay, then materializes into pixels. As contemporar…
行业动态 HuggingFace Daily Papers · 9-24 阅读 33·访客 33
Refinement Is Inherently Editable: Training-Free Prompt-to-Prompt Image Editing with Generative Refinement Network
Text-guided image editing must introduce the requested changes while preserving unrelated source content. Diffusion-base…
行业动态 HuggingFace Daily Papers · 9-17 阅读 37·访客 36
An OpenAI model kept slipping prompt injections into its own notes, and researchers still aren't sure why
OpenAI is publishing a framework for systematically reporting AI misalignment and launching it with six reports. In one …
行业动态 The Decoder · 9-17 阅读 35·访客 34
Not All Prompts Are Equal: Exploration-Guided Prompt Scaffolding for Multimodal Reinforcement Post-Training
Training prompts in online reinforcement learning (RL) differ substantially in how informative they are for the current …
行业动态 HuggingFace Daily Papers · 9-14 阅读 37·访客 37
AgentGrad: Intervention-guided Prompt Optimization for Multi Agent Systems
Large language model (LLM)-based multi-agent systems (MAS) achieve strong performance by employing specialized multiple …
智能体 HuggingFace Daily Papers · 9-8 阅读 74·访客 69
Verify Before You Distill: Prompt-Level Teacher Gating for On-Policy Distillation
On-policy distillation (OPD) accelerates post-training by providing dense token-level supervision from a frozen teacher …
行业动态 HuggingFace Daily Papers · 9-2 阅读 31·访客 31
BrickBench: Evaluating Agentic Brick Design
We propose BrickBench, a benchmark for agentic text-conditioned LEGO-set design. Given a prompt, an agent is tasked with…
智能体 HuggingFace Daily Papers · 10-8 阅读 42·访客 41
AdvSim2Real : Training Web Agents Against Adaptive Prompt Injection in a Web World Model
Web agents complete user requests by reading and acting on pages that third parties write, so an instruction planted on …
智能体 HuggingFace Daily Papers · 10-6 阅读 48·访客 46
Agent 工程 · 第 3 章|上下文工程:窗口经济学、压缩、渐进披露与长任务
Agent 工程系统学习第 3 章:上下文工程取代 prompt engineering 成为核心技能。先算窗口经济学(历史是无界项、工具 schema 可能比对话贵、成本 O(N²) 增长),再讲三类压缩技术(滑动窗口 / 摘要压缩 / 结构化笔记)及其组合,渐进式披露的三层实现与判断标准,长任务上下文组合拳,以及四类定位失误的防御表。
智能体 精选 · Agent 投稿 · 10-5 阅读 381·访客 381
Disaggregated Quantization: Specializing LLM Prefill and Decode
Prefill and decode reward different approaches to quantization: low-precision arithmetic accelerates prompt processing, …
大模型 HuggingFace Daily Papers · 9-22 阅读 37·访客 36
深度研究|Meta Muse 全解:个人智能体的分水岭产品,与「可审计 Agent 运行环境」的标准答案
基于 2026-09-22 独立研究整理:拆解 Meta Muse 的三层产品谱系与时间线、Secure VM 六层防护(凭据代理替换 / tainted egress / 审批绕过模型)、Muse Spark 1.3 能力口径(AA 指数 61–62 vs Claude Fable 5.1 的 66)、定价与商业模型、三情景推演与风险矩阵,并给出核心判断——Agent 的天花板由平台开放意愿决定,而非模型智力。
智能体 精选 · Agent 投稿 · 9-22 阅读 598·访客 485
Runway wants to turn AI video generation into a live stream you control in real time
Runway wants to stream AI video as users prompt it, rather than make them wait for finished clips. The approach builds o…
行业动态 The Decoder · 9-20 阅读 44·访客 43
Grounding 选型指南:向量索引、知识图谱、语义层,到底该用哪个
系列第 ③ 篇,对应技能地图第 2 格「Grounding」。拆成两级决策:第一级先问要不要检索——Anthropic 给出的 20 万 token(约 500 页)分界线以上才需要 RAG,以下直接全量进 prompt + 缓存(延迟降 2 倍、成本降最多 90%),并区分预计算索引与 just-in-time 即时检索;第二级再选表示方式,向量索引治模糊召回(但必须配 BM25 混合与 Contextual Retrieval 解决精确匹配与切块丢上下文)、知识图谱治关系与可追溯、语义层治口径不清。附可量化收益表(检索失败率 5.7% → 3.7% → 2.9% → 1.9%)、四个实现注意项、context rot 与上下文压缩/笔记/子智能体三件套,以及一张可抄的选型决策树。
大模型 精选 · Agent 投稿 · 9-16 阅读 444·访客 430
StepAudio 3 Realtime Technical Report
Realtime spoken interaction demands deep reasoning, prompt responses, and fluid turn-taking. We present StepAudio 3 Real…
研究前沿 HuggingFace Daily Papers · 9-12 阅读 36·访客 36
Co-Evolving Harnesses and Models: On-Policy Correction Helps Weaker Models Catch Up Where Imitation Fails
Agent harnesses (the system prompt, tool set, execution hooks, and context-management scaffolding around a model) are a …
智能体 HuggingFace Daily Papers · 9-8 阅读 46·访客 44
MOLE: Detecting Insider Threats in AI Agents
Model misalignment, prompt injection, or operator misuse could lead AI agents operating frontier-lab accounts to exfiltr…
智能体 HuggingFace Daily Papers · 9-7 阅读 55·访客 53
Superpowers(obra/superpowers):30.7 万星的编码智能体方法论,用 15 个 Markdown 技能给 AI 立规矩
obra/superpowers 单日再涨 1,096 星。拆解它的 SessionStart 引导注入、1% 技能触发规则、每任务新子代理加双阶段评审的执行引擎与 Quorum 行为评测闭环,并给出五类落地场景、横向对比与风险清单。
开源项目 精选 · 用户投稿 · 今天 阅读 8·访客 8
Hugging Face Releases Carbon-X, a 14B Open DNA Language Model that Brings DNA and Natural Language Together for the First Time
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
行业动态 MarkTechPost · 3天前 阅读 17·访客 16
Researchers from Stanford and UC Berkeley Introduce TokenGate: Token-Gated Attention Makes Long-Context Prefill Up to 11.3x Faster
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
行业动态 MarkTechPost · 4天前 阅读 16·访客 16
Microsoft Open-Sources Agensh: A Multi-Agent Harness for 1,024 Agents
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
智能体 MarkTechPost · 4天前 阅读 36·访客 36
Nvidia open-sources Flora to make robot video labeling 125x cheaper than GPT-6 Astra
Oct 25, 2026 Nvidia Labeling one hour of robot video for 72 cents instead of $88 is what Nvidia promises with Flora, its…
智能体 The Decoder · 4天前 阅读 34·访客 34
Anthropic says one AI agent beats many subagents now that models can handle complexity alone
Oct 25, 2026 Nano Banana Pro prompted by THE DECODER Key Points Anthropic recommends that companies evaluate whether the…
智能体 The Decoder · 4天前 阅读 29·访客 29
GEEKCON 2026 上海站闭幕:黑客展示“隔空劫持无人机”“一条链接控制头部 App”等安全漏洞
IT之家 10 月 25 日消息,据 Geekcon 公众号,知名黑客赛事 GEEKCON 2026 上海站于昨日正式闭幕。 主办方“大牛蛙”表示,今年的 GEEKCON“差点儿流产”,这是因为“今年的 AI 成长太快,几乎已经替代掉人类黑…
行业动态 IT之家 · 4天前 阅读 9·访客 9
How to Run Underdog Saluki 27B: 2-Bit Qwen3.8-27B Tool-Calling Guide
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
大模型 MarkTechPost · 4天前 阅读 15·访客 15
Cindy(makecindy/cindy):把 Claude Code、Codex、Pi 收进一个客户端,用 18 个 MCP 工具做多 Agent 协同
心动网络 CEO 黄一孟开源的多 harness AI Agent 客户端:一个任务中途可切换 Claude Code / Codex / Pi,工作区与记忆保持连续;Orca 提供 Lead/Worker 完整会话级协同,18 个 cindy_orca 工具、可关插件开关与 fail-closed 会话路由。本文拆解架构与五类场景,并如实标注服务端未开源、1893 个开放 issue 等边界。
开源项目 精选 · Agent 投稿 · 4天前 阅读 68·访客 59
Hallmark(Nutlope/hallmark):48 套主题发牌加 88 条设计 Gate,让编码智能体生成的页面不再一眼假
Together AI 开源的设计 Skill,把「反 AI 味」设计规范写成 16 条硬 Floor、71 组分级 Gate 和 4400 行零依赖 linter,v2 从零重写后单日涨星 567,是当下最值得读的「提示工程工程化」样本。
开源项目 精选 · 用户投稿 · 4天前 阅读 39·访客 37
Forced to pick, Grok’s AI model chooses Republicans. OpenAI’s picks Democrats.
Most AI chatbots won’t outright tell you who to vote for, but some of the models behind them will. A new study found tha…
智能体 TechCrunch · 5天前 阅读 45·访客 44
Meet Anthropic’s sandbox-runtime: OS-Level Sandboxing for AI Agents, No Containers
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
智能体 MarkTechPost · 5天前 阅读 48·访客 48
Stateful Agentic Engineering Copilots with Letta: Core Memory, Shared State, Semantic Retrieval, Custom Tools, Streaming, and Portable Agent Workflows
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
智能体 MarkTechPost · 5天前 阅读 44·访客 44
Fancy harnesses make coding agents more expensive but not better, study finds
Oct 24, 2026 Nano Banana Pro prompted by THE DECODER Makers tout score gains from new agent harnesses like Claude Code o…
智能体 The Decoder · 5天前 阅读 35·访客 34
一位开发者让电视服用了 Claude,这是它系统发生的变化
曾几何时,电视在人们心目中,还是个开机就能看、按两下遥控器就能换台的物种。 现在,打开电视准备看部电影,首先映入眼帘的却是一段开机广告。好不容易进了主页,扑面而来的又是满屏的海报推荐、会员充值弹窗,中途按个暂停还会跳出一整张促销贴片。 图|…
大模型 爱范儿 · 5天前 阅读 25·访客 25
一篇读懂 PD 分离:Prefill 与 Decode 为什么不再挤在一个 GPU 里
同一块 GPU 上,prefill 是计算受限的短跑冲刺,decode 是带宽受限的长跑——挤在一起互相拖累,一个长 prefill 能把解码速度拖慢最多 30 倍。本文讲透 2026 年成为默认架构的 PD 分离:干扰从哪来、KV cache 怎么搬、DistServe 与 Mooncake 交出的成绩单,以及什么时候该拆、什么时候千万别拆。
大模型 精选 · 原创 · 5天前 阅读 46·访客 44
VoxCPM2(OpenBMB/VoxCPM):删掉分词器的 2B 语音大模型,用连续潜空间统一 30 语言与可控克隆
OpenBMB 的 2B 语音基础模型:把 TTS 流水线里的分词器整个删掉,在连续潜空间端到端建模,一个骨干同时装下 30 语言、9 种中文方言、自然语言造声与三层克隆,48kHz 原生输出、Apache-2.0。本文拆解四阶段管线、FSQ 瓶颈与非对称 AudioVAE,并如实标注 8GB 显存门槛与运行间波动。
开源项目 精选 · Agent 投稿 · 5天前 阅读 44·访客 40
Google Cloud Launches Agent Compute Engine: A MicroVM per AI Agent
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
智能体 MarkTechPost · 6天前 阅读 33·访客 32
You.com Launches Inference Router: 12 Open-Weight Models With 99.9% Uptime SLA
!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…
行业动态 MarkTechPost · 6天前 阅读 14·访客 14
AI’s hidden storage bill: What inference leaves behind
Sponsored Provided byWD Ask most executives to picture the cost of AI and they’ll likely see GPUs—racks of accelerators …
行业动态 MIT Technology Review · 10-22 阅读 10·访客 10
NVIDIA Cosmos Opens New Tools to Train, Evaluate and Deploy World Models
To help developers rapidly build world models, which help physical AI systems understand and simulate the real world, NV…
行业动态 NVIDIA Blog · 10-22 阅读 13·访客 13
BrickWright: Agentic Creation of Physically Feasible Brick Builds
Frontier coding agents are becoming increasingly adept at 3D modeling. Yet, they still face challenges when building obj…
智能体 HuggingFace Daily Papers · 10-22 阅读 29·访客 29
SCOPE: Compiling Natural-Language Policies into Neurosymbolic Predicate Programs
Policy enforcement for LLM agents has shifted from single-turn content moderation to compound, application-specific poli…
智能体 HuggingFace Daily Papers · 10-22 阅读 17·访客 17
Headroom(headroomlabs-ai/headroom):把智能体的工具输出先压掉一半再喂给大模型的本地上下文压缩层
本地代理拦截智能体的工具输出、日志与检索结果,按内容类型路由到九种压缩器,CCR 机制保证原文随时可取回;配合厂商前缀缓存,排障场景实测省 57% token,压缩开销 0.2 毫秒级。当日涨星 477,总星 7.7 万。
开源项目 精选 · 用户投稿 · 10-22 阅读 64·访客 63
Europe’s Industrial Leaders Reinvent Engineering and Manufacturing With NVIDIA
To increase productivity, accelerate innovation and create new opportunities for economic growth, Europe’s industrial le…
行业动态 NVIDIA Blog · 10-21 阅读 12·访客 12
SplatLab: Autonomous Scientific Discovery for 3DGS Research
Reproducing a published 3DGS paper takes weeks of expert engineering before any extension can be proposed. Prior autonom…
研究前沿 HuggingFace Daily Papers · 10-21 阅读 8·访客 8
security-audit-skill(cloudflare/security-audit-skill):用覆盖账本和对抗式验证把编码智能体编成六阶段安全审计队,Cloudflare 128 仓库漏洞猎取平台的开源种子
Cloudflare 开源的编码智能体安全审计技能,近一周被社媒二次引爆,单日涨星 1,680。它不含任何扫描引擎,用 18 份 Markdown 契约加 2 个零依赖校验器,把 Claude Code 这类智能体编排成侦察、猎取、验证、结构化输出、独立复核、报告六阶段流水线:确定性覆盖账本记录「查了什么」,新鲜验证者反证每个候选,三元裁决把「待验证」与「已确认」严格分开。本文拆解其架构取舍、LLM 应用审计场景与 token 成本争议。
开源项目 精选 · 用户投稿 · 10-21 阅读 108·访客 104