KK Blog

Menu

All Tags

#证明覆盖 1 #软件安全 1 #Proof Coverage 1 #OSS-Fuzz 2 #RISC-V 2 #Software Security 1 #验证 1 #反诈 1 #金融安全 1 #验证码 1 #社会工程 1 #独立渠道核验 1 #冒充银行 1 #电话诈骗 1 #Vishing 1 #Financial Security 1 #Verification Codes 1 #Social Engineering 1 #Bank Fraud 1 #Imposter Scams 1 #推理评测 1 #批次不变性 1 #确定性推理 1 #Inference Evaluation 1 #Batch Invariance 1 #Deterministic Inference 1 #DeepInfer 2 #模型可见内容 1 #RAG安全 1 #间接提示注入 1 #检索血缘 1 #Model-Visible Content 1 #Indirect Prompt Injection 1 #Retrieval Provenance 1 #行为等价 1 #软件评测 1 #COBOL现代化 1 #低资源编程语言 1 #遗留系统迁移 1 #MultiLangBench 2 #Behavioral Equivalence 1 #Software Evaluation 1 #COBOL Modernization 1 #Low-Resource Programming Languages 1 #Legacy Code Migration 1 #可靠性 1 #Agent路由 1 #AI Agent架构 1 #薄编排器 1 #Context Isolation 1 #Reliability 1 #Agent Routing 1 #AI Agent Architecture 1 #Thin Orchestrator 1 #本地大模型 1 #工具调用 1 #混合推理 1 #潜空间推理 1 #过拟合 1 #基准泄漏 1 #自我改进Agent 1 #M5 Ultra 2 #Mac Studio 2 #Tool Use 1 #Local LLM 1 #MoE 2 #Hybrid Reasoning 1 #Muse Spark Open 2 #Latent Reasoning 1 #HybridAgent 2 #AI Evaluation 1 #Overfitting 1 #Benchmark Leakage 1 #Self-Improving Agents 1 #Harness² 2 #验证吞吐 1 #AI水印 1 #C2PA 2 #Verification Throughput 1 #AI Watermarking 1 #软件质量 1 #Software Quality 1 #算法设计 1 #量子计算 1 #Toffoli门 1 #AI科研 1 #量子电路 1 #Algorithm Design 1 #Quantum Computing 1 #Toffoli Gates 1 #AI for Science 1 #Quantum Circuits 1 #安全研究 1 #可复现性 2 #AI科学发现 1 #密码学 1 #NIST曲线 1 #Security Research 1 #AI Discovery 1 #Cryptography 1 #NIST Curves 1 #ECDSA 2 #人力规划 1 #岗位自动化 1 #AI基准 1 #Mercor 2 #Workforce Planning 1 #Job Automation 1 #AI Benchmarks 1 #Meta Forum 2 #Source Provenance 1 #GEO 1 #RAG Security 2 #Citation Security 1 #AI Search 1 #Runtime Architecture 1 #Rust 1 #Third-Party Assessment 1 #运行时监控 1 #模型供应链 1 #密码学后门 1 #AI安全审计 1 #能力诱导 1 #模型隐藏能力 1 #Runtime Monitoring 1 #Model Supply Chain 1 #Cryptographic Backdoors 1 #AI Safety Audits 1 #Capability Elicitation 1 #Hidden AI Capabilities 1 #版本控制 1 #开发者基础设施 1 #源码管理 1 #Version Control 1 #Developer Infrastructure 1 #Monorepo 2 #Source Control 1 #Anthropic Scion 2 #误伤 1 #滥用检测 1 #推理保护 1 #模型蒸馏 1 #Frontier Models 2 #False Positives 1 #Abuse Detection 1 #Reasoning Protection 1 #Model Distillation 1 #负载卸载 1 #容量规划 1 #弹性推理 1 #GPU利用率 1 #AI推理 1 #Load Shedding 1 #Capacity Planning 1 #Elastic Inference 1 #GPU Utilization 1 #AI Inference 1 #遏制 1 #AI蠕虫 1 #IoT安全 1 #自主复制 1 #AI自我复制 1 #Containment 1 #AI Worms 1 #IoT Security 1 #Autonomous Replication 1 #AI Self-Replication 1 #Agent上下文压缩 1 #AI Agent Context Compression 1 #有状态Agent评测 1 #pass^20 1 #pass@20 2 #Stateful Agent Benchmark 1 #AI Agent Reliability 1 #ThinkingBox-Bench 2 #ThinkingBox 2 #可复现安全 1 #威胁建模 1 #AI漏洞报告 1 #Reproducible Security 1 #Threat Modeling 1 #AI Vulnerability Reports 1 #Anthropic OSS Scanner 2 #UI编译器 1 #流式UI 1 #生成式界面 1 #UI Compiler 1 #Streaming UI 1 #Generative UI 1 #Intelligent UI 2 #审计日志 1 #RFC 8915 2 #Chrony 2 #NTP 2 #NTS 2 #Network Time Security 2 #硬件兼容 1 #华为昇腾 1 #Hardware Portability 1 #FlashMLA 2 #DeepEP 2 #DeepGEMM 2 #Huawei Ascend 1 #GPU性能 1 #投机解码 1 #GPU Performance 1 #Speculative Decoding 1 #Kimi K2.6 2 #科研AI 1 #响应面 1 #实验设计 1 #AI研究Agent 1 #Response Surface 1 #Experimental Design 1 #WhatWorkedBench 2 #AI Research Agents 1 #AI评测 2 #安全数据 1 #对抗训练 1 #自动化红队 1 #AI Evals 1 #Security Data 1 #Adversarial Training 1 #Automated Red Teaming 1 #数据泄漏 1 #隐藏测试 1 #边界条件 1 #程序修复 1 #Coding Agent评测 1 #工具输出 1 #检索增强 1 #来源验证 1 #RAG评测 1 #LLM权威偏差 1 #Data Leakage 1 #Hidden Tests 1 #Boundary Cases 1 #Program Repair 1 #Coding Agent Evaluation 1 #Failure Map 2 #NeurIPS 2026 4 #Tool Outputs 1 #Source Verification 1 #RAG Evaluation 1 #LLM Authority Bias 1 #Research Agents 1 #BootLoops 1 #Claude-shaped science 1 #Cyber Defense 1 #Frontier Safety Framework 1 #Fairwind Program 1 #Phased Access 1 #Gemini 4 Argon 1 #O*NET 3 #Robotics Economics 1 #Automation Forecast 1 #Robot Adoption 1 #Robot Exposure 1 #Australia 1 #Authorization 1 #Incident Evidence 1 #Government Systems 1 #Web追踪 1 #移动隐私 1 #聊天机器人追踪 1 #对话式AI隐私 1 #Web Tracking 1 #Mobile Privacy 1 #Chatbot Tracking 1 #Conversational AI Privacy 1 #AI滥用 1 #网络能力 1 #网络安全AI 1 #NIST CAISI 2 #AI Misuse 1 #Cyber Capability 1 #Cybersecurity AI 1 #GLM-5.3 2 #AI审计 1 #训练治理 1 #AI安全论证 1 #前沿AI 1 #AI Audit 1 #Misalignment 1 #Training Governance 1 #AI Safety Case 1 #全文检索 1 #对话历史 1 #会话留存 1 #claude-mem 2 #Funes 2 #Full Text Search 1 #Agent Memory 1 #Conversation History 1 #Session Retention 1 #DNS隧道 1 #网络隔离 1 #DNS出口 1 #DNS Tunneling 1 #Network Isolation 1 #Sandbox Containment 1 #DNS Egress 1 #引证核验 1 #事项治理 1 #法律检索 1 #AI审计轨迹 1 #法律AI 1 #Citation Verification 1 #Matter-Centric Governance 1 #Legal Research 1 #AI Audit Trail 1 #Astra for Law 2 #Legal AI 1 #可穿戴AI 1 #数据删除 1 #可信执行环境 1 #机密计算 1 #隐私 1 #AI眼镜 1 #Wearable AI 1 #Data Deletion 1 #Trusted Execution Environment 1 #Confidential Computing 1 #AI Glasses 1 #风险指标 1 #临床AI 1 #心理健康AI 1 #Risk Metrics 1 #Clinical AI 1 #Mental Health AI 1 #MentalHealthBench 2 #代码隐私 1 #数据流审计 2 #仓库安全 1 #Source Code Privacy 1 #ZCode 2 #Data Flow Audit 2 #Repository Security 1 #LLM可观测性 1 #API成本 1 #显式断点 1 #缓存诊断 1 #Prompt缓存 1 #LLM Observability 1 #API Cost 1 #Explicit Breakpoints 1 #Cache Diagnostics 1 #OpenAI API 2 #Prompt Caching 1 #GPT-6 4 #系统工程 1 #约束求解 1 #资源分配 1 #混合整数规划 1 #局部搜索 1 #分配优化 1 #Systems Engineering 1 #Constraint Solving 1 #Resource Allocation 1 #Mixed Integer Programming 1 #Local Search 1 #Assignment Optimization 1 #Meta Rebalancer 2 #科学传播 1 #科研治理 1 #数学验证 1 #独立审查 1 #Scientific Communication 1 #Research Governance 1 #Mathematical Verification 1 #Independent Review 1 #数据溯源 1 #机构记忆 1 #AI Agent记忆 1 #Knowledge Graph 1 #Data Provenance 1 #Context Graph 2 #Institutional Memory 1 #Amdahl定律 1 #测试影响分析 1 #技术债 1 #软件重写 1 #Amdahl's Law 1 #Test Impact Analysis 1 #Technical Debt 1 #Software Rewrite 2 #恢复协议 1 #独立验证 1 #前沿模型 2 #AI事件响应 1 #Recovery Protocol 1 #Independent Verification 2 #Frontier AI 2 #AI Incident Response 1 #AI Kill Switch 2 #生物安全 1 #AI访问控制 1 #湿实验验证 1 #蛋白质设计 1 #生命科学AI 1 #Biosafety 1 #Claude Science 2 #AI Access Controls 1 #Wet Lab Validation 1 #Protein Design 1 #Life Science AI 1 #独立评估 1 #可复现证据 1 #AI事件报告 1 #模型失配 2 #Independent Evaluation 2 #Reproducible Evidence 1 #AI Incident Reporting 1 #Model Misalignment 1 #LLM架构 1 #主机内存卸载 1 #稀疏模型 1 #混合专家 1 #条件记忆 1 #LLM Architecture 1 #Host Memory Offload 1 #Sparse Models 1 #Mixture of Experts 1 #Conditional Memory 1 #DeepSeek Engram 2 #轮次交互 1 #语音AI延迟 1 #语音Agent评测 1 #Voice Agent Evaluation 1 #Live API 2 #Extended Thinking 2 #Turn Taking 1 #Voice AI Latency 1 #Voice Agent Benchmark 1 #Gemini 3.8 Live 2 #智能家居可靠性 1 #unavailable状态 1 #unknown状态 1 #传感器过期 1 #API自动化 1 #Smart Home Reliability 1 #Template Sensor 2 #Unavailable State 1 #Unknown State 1 #Stale Sensor 1 #API Automation 1 #AI供应链 2 #Prompt路由 1 #LLM API代理 1 #LLM中转站安全 1 #AI Supply Chain 2 #Prompt Routing 1 #LLM API Proxy 1 #LLM Relay Security 1 #滑动窗口注意力 1 #SWA有界重放 1 #Sliding Window Attention 1 #SWA Bounded Replay 1 #DeepSeek V4.1 4 #可复现仪表盘 1 #SQL溯源 1 #数据血缘 1 #可验证分析 1 #Reproducible Dashboard 1 #SQL Provenance 1 #Data Lineage 1 #Verifiable Analytics 1 #ChatGPT Data Agent 2 #Agent状态 1 #Agent迁移 1 #托管式Harness 1 #Codex Harness 2 #Agent State 1 #Agent Migration 1 #Managed Agent Harness 1 #OpenAI Agents API 2 #Bot管理 1 #AI爬虫 1 #Bot Management 1 #robots.txt 2 #Cloudflare BotBase 2 #AI Crawlers 1 #推理状态 1 #交互式Agent 1 #人工接管 1 #科学AI 1 #测量合同 1 #量子校准 1 #实验室Agent 1 #事件分析 1 #GET请求 1 #只读网络 1 #OpenAI Agent 1 #Inference State 1 #Interactive Agents 1 #AsyncReasoning 2 #Scientific AI 3 #Measurement Contract 1 #GPT-5.6 Sol 2 #Quantum Calibration 1 #AI Lab Agents 1 #Incident Analysis 1 #GET Requests 1 #Read-Only Web 1 #OpenAI Agents 1 #DseWiki 2 #上下文压缩 1 #模型评测 1 #Context Compaction 3 #Provider Adapter 2 #ARC-AGI-3 2 #GPT-6 Astra 2 #AI Mathematics 2 #Prove2Me 1 #Fermat's Last Theorem 1 #主动式AI 1 #检查点 1 #Agent权限 1 #长任务Agent 2 #持久Agent 1 #Proactive AI 1 #Checkpointing 1 #AI Agent Permissions 1 #Long-Running Agents 2 #Persistent AI Agents 1 #知识溯源 1 #专家反馈 1 #企业知识管理 1 #组织第二大脑 1 #RAG 4 #Knowledge Provenance 1 #Expert Feedback 1 #Enterprise Knowledge Management 1 #Organizational Second Brain 1 #回滚 2 #可恢复性 1 #自我修改Agent 1 #Rollback 2 #Recoverability 1 #Self-Modifying Agents 1 #EvoUndo 2 #Model Evaluation 1 #Terminal-Bench 2 #Claude Mythos 5.1 2 #Claude Fable 5.1 2 #Privacy 2 #Algorithmic Auditing 1 #Advertising Audit 1 #Answer Independence 1 #ChatGPT Ads 1 #幻觉 1 #推理 1 #知识提取 1 #参数知识 1 #LLM事实性 1 #Hallucination 1 #Reasoning 1 #Retrieval 2 #LLM Evaluation 4 #WikiProfile 2 #Parametric Knowledge 1 #LLM Factuality 1 #工作负载身份 1 #MCP授权 1 #Agent身份 2 #AI Sovereignty 1 #Continual Learning 1 #Build vs Buy 1 #Domain-Specific LLM 1 #Thomson 1.0 1 #Thomson Reuters 1 #DPoP 2 #ID-JAG 2 #Workload Identity 1 #OAuth 2.1 2 #MCP Authorization 1 #Agent Identity 2 #奖励作弊 1 #留出测试 1 #AI安全评测 1 #AI对齐 1 #自动对齐研究员 1 #Petri 2 #Held-Out Evaluation 1 #AI Safety Benchmarks 1 #AI Alignment 1 #Automated Alignment Researchers 1 #加州律师协会 1 #审计链 1 #高风险认证 1 #律师考试 1 #AI披露 1 #加州AB1651 1 #State Bar of California 1 #Audit Trail 2 #High-Stakes Certification 1 #Bar Exam 1 #AI Disclosure 1 #California AB 1651 1 #AI控制面 1 #工作区管理 1 #AI Control Plane 1 #Workspace Administration 1 #RBAC 2 #ChatGPT Work 4 #OpenAI Admin Plugin 2 #本地AI 1 #推理强度 1 #量化 1 #Local AI 1 #Reasoning Effort 1 #KV Cache 8 #SGLang 8 #Quantization 1 #LLM Benchmarking 2 #Qwen3.8-27B 2 #Agent推理 1 #能效 1 #基准审计 1 #AI推理芯片 1 #Agentic Inference 1 #Power Efficiency 1 #Benchmark Audit 1 #InferenceX 2 #AI Inference Chip 1 #OpenAI Jalapeño 2 #AI加速器 1 #Meta AI基础设施 1 #AI网络 1 #AI Accelerators 1 #HCCL 2 #Meta AI Infrastructure 1 #AI Networking 1 #MetaRoCE 2 #MTIA 300 2 #LLM Agent 2 #程序性记忆 1 #Agent Skill 1 #LLM Agents 2 #Agent Evals 2 #Procedural Memory 1 #Agent运行时 2 #工作流编排 1 #持久化执行 1 #Platform Units 2 #Temporal 2 #Workflow Orchestration 1 #Durable Execution 1 #UiPath Maestro Flow 2 #平台中立性 1 #AI网关 1 #Token计费 1 #AI财务控制面 1 #模型路由 1 #Platform Neutrality 1 #AI Gateway 1 #Token Billing 1 #AI Financial Control Plane 1 #AI Model Routing 1 #OpenRouter 2 #Stripe 2 #App Store Review 1 #In-Process Runtime 1 #Android Sandbox 1 #iOS Sandbox 1 #PhoneBuddySDK 1 #Mobile AI Agents 1 #信用风险 1 #项目融资 1 #剩余价值担保 1 #数据中心融资 1 #Credit Risk 1 #Project Finance 1 #Residual Value Guarantee 1 #Data Center Finance 1 #地理空间分析 1 #证据链 1 #图像地理定位 1 #OpenStreetMap 2 #Geospatial Analysis 1 #CUDA 2 #Evidence Chain 1 #OSINT 2 #Image Geolocation 1 #Agent架构 1 #任务复杂度 1 #AI成本优化 1 #小语言模型 1 #LLM模型路由 1 #软件维护 1 #Agent指标 1 #Pull Request审查 1 #AI编程Agent 3 #Anthropic风险报告 1 #控制面 1 #安全分类器 1 #安全日志 1 #AI安全控制 1 #Task Complexity 1 #Model Router 2 #AI Cost Optimization 1 #Small Language Models 1 #LLM Model Routing 1 #Software Maintenance 1 #CI 4 #Agent Metrics 1 #Pull Request Review 1 #Claude Code Routines 2 #Anthropic Risk Report 1 #Control Plane 1 #Safety Classifiers 1 #Security Logging 1 #AI Safety Controls 1 #置信度路由 1 #边缘AI 1 #隐私架构 1 #家庭自动化 1 #本地VLM 1 #Home Assistant 4 #Confidence Routing 1 #LFM2.5-VL-3B 2 #Privacy Architecture 1 #Home Automation 1 #Local VLM 1 #自回归解码 1 #LLM推理 5 #并行解码 1 #扩散语言模型 1 #Autoregressive Decoding 1 #AIME 2026 2 #vLLM 2 #LLM Inference 5 #Parallel Decoding 1 #Diffusion LLM 1 #DiffusionGemma 2 #Agent协调 1 #委托保持 1 #多Agent系统 2 #Human Oversight 2 #Agent Coordination 1 #Goal Preservation 1 #AI Agent Governance 1 #ITT 1 #劳动力政策 1 #Sector Program 1 #随机对照试验 1 #AI岗位替代 1 #职业再培训 1 #Intent-to-Treat 1 #Workforce Policy 1 #Sector Programs 1 #Randomized Trials 1 #AI Job Displacement 1 #Worker Retraining 1 #差分隐私 1 #透明账本 1 #机密联邦分析 1 #端侧机器学习 1 #Differential Privacy 1 #TEE 2 #Transparency Ledger 1 #Confidential Federated Analytics 1 #On-Device ML 1 #WhatsApp Scam Alert 2 #生成式 AI 1 #德国 1 #记忆化 1 #模型权重 1 #著作权 1 #AI 音乐 1 #Germany 1 #GEMA 2 #Memorization 1 #Model Weights 1 #Suno 2 #软件验证 1 #go vet 2 #go test 2 #gofmt 2 #Software Verification 1 #Agent Harness 6 #Go 2 #上下文绑定 1 #API安全 2 #Agent日志 1 #加密推理 1 #Context Binding 1 #API Security 2 #Agent Logs 1 #Encrypted Reasoning 1 #24GB显存 1 #本地Agent 1 #Edge AI 2 #DFlash 2 #24GB VRAM 1 #Local AI Agent 1 #Meta AI 4 #Muse Glimmer 2 #黎曼假设 1 #黎曼ζ函数 1 #AI for Mathematics 1 #Lean 4 3 #Riemann Hypothesis 1 #Riemann Zeta 1 #Agent Runtime 3 #Idempotency 1 #Background Agents 1 #Crash Recovery 1 #Event Sourcing 1 #Meta Muse Code 1 #科学验证 1 #后量子密码 1 #AI密码分析 1 #Scientific Verification 1 #Post-Quantum Cryptography 1 #AES 2 #HAWK 2 #AI Cryptanalysis 1 #浏览器自动化 1 #V8 Isolate 1 #Agent浏览器 1 #WebAssembly 2 #Browser Automation 1 #Chromium 2 #Browser Run 2 #V8 Isolates 1 #Agent Browser 1 #Cloudflare Kitesurf 2 #网络安全评测 1 #AI网络安全 1 #Critical网络能力 1 #Cyber Evaluations 1 #Agentic Coding 2 #AI Cybersecurity 1 #Preparedness Framework 2 #Critical Cyber Capabilities 1 #OpenAI Astra 2 #在线推理 1 #用户序列建模 1 #Scaling Law 1 #多阶段推荐系统 1 #Meta广告排序 1 #Online Inference 1 #User Sequence Modeling 1 #Scaling Laws 1 #LLaTTE 2 #Multi-Stage Recommender Systems 1 #Meta Ads Ranking 1 #端点安全 1 #运行时阻断 1 #安全钩子 1 #AI Agent安全 2 #可复现研究 1 #模拟校准 1 #AI验证 2 #科学计算 1 #国家实验室 1 #科学Agent 1 #Numbat 2 #Endpoint Security 1 #CEL 2 #Runtime Enforcement 1 #Security Hooks 1 #AI Agent Security 4 #Reproducibility 3 #Simulation-Based Calibration 1 #AI Verification 2 #Scientific Computing 1 #National Laboratories 1 #Scientific AI Agents 1 #Jagged Tensor 1 #GPU效率 1 #LLM训练 1 #推荐系统 3 #Jagged Tensors 1 #MFU 2 #GPU Efficiency 1 #LLM Training 1 #Recommendation Systems 3 #Meta GEM 2 #Privacy Verification 1 #SHA-256 1 #Consumer Privacy 1 #Data Removal 1 #Privacy Workflow 1 #Delete Act 1 #Data Broker Deletion 1 #California DROP 1 #开发者体验 1 #动态任务评测 1 #产品AI可用性 1 #Developer Experience 1 #AGENTS.md 2 #Dynamic Task Evals 1 #Product Evaluation 1 #AI Usability 1 #AI Agent Evals 1 #漏洞管理 1 #Security Agent 1 #SQLite安全 1 #自动修复 1 #漏洞复现 1 #CVE核验 1 #AI垃圾CVE 1 #Vulnerability Management 1 #ASan 2 #Security Agents 1 #SQLite Security 1 #Auto-Remediation 1 #Vulnerability Reproduction 1 #CVE Verification 1 #AI Slop CVE 1 #二手硬盘 1 #数据隐私 1 #密码学擦除 1 #介质清除 1 #SSD安全清除 1 #SSD Resale 1 #Data Privacy 1 #Cryptographic Erase 1 #NVMe Sanitize 2 #NIST SP 800-88 2 #Media Sanitization 1 #SSD Secure Erase 1 #可验证AI 1 #语义中间层 1 #AI数据可视化 1 #Verifiable AI 1 #Chart.js 2 #ECharts 2 #Vega-Lite 2 #Semantic IR 1 #AI Data Visualization 1 #Flint Chart 2 #可观测性 2 #事件合同 1 #语音转文字 1 #Voice Agent 1 #实时转写 1 #Observability 2 #Event Contracts 1 #Speech to Text 1 #OpenAI Realtime API 2 #GPT Live Transcribe 2 #Voice Agents 1 #Realtime Transcription 1 #长上下文 2 #提交门控 1 #AI Agent规则 1 #Long Context 2 #HANDBOOK.md 2 #Commit Gates 1 #AI Agent Policy 1 #产品分析 1 #零售AI 1 #收入归因 1 #对话式AI 1 #语音Agent分析 1 #Product Analytics 1 #AI ROI 2 #GPT-Realtime 2 #Retail AI 1 #Revenue Attribution 1 #Conversational AI 1 #Voice Agent Analytics 1 #内容溯源 2 #AI部署者 1 #AI提供者 1 #AI合规 1 #AI透明度 1 #欧盟AI法案 1 #Content Provenance 2 #AI Deployers 1 #AI Providers 1 #AI Transparency 1 #Article 50 2 #N-day风险 1 #安全自动化 1 #补丁管理 1 #漏洞分诊 2 #Chrome安全 1 #AI漏洞管理 1 #N-day Risk 1 #Security Automation 1 #Patch Management 1 #Vulnerability Triage 2 #Chrome Security 1 #AI Vulnerability Management 1 #信息流控制 1 #上下文隔离 2 #提示注入 2 #文档携带型AI蠕虫 1 #Information Flow Control 1 #Microsoft 365 2 #Context Quarantine 1 #Prompt Injection 4 #Microsoft Copilot 2 #Document-Borne AI Worm 1 #事件响应 2 #轨迹监测 1 #长任务AI安全 1 #Incident Response 2 #Trajectory Monitoring 1 #Long-Horizon AI Safety 1 #思维链 3 #CloudWatch 2 #Chain-of-Thought 3 #OpenTelemetry 2 #AI Agent Observability 1 #模型选型 2 #Coding Agent 7 #AI Agent评测 3 #Coding Agents 6 #AI Agent Evaluation 1 #AI Agent Benchmark 2 #恢复演练 1 #灾难恢复 1 #备份 2 #RTO 2 #RPO 2 #Git 4 #BorgBackup 2 #Restore Testing 1 #Disaster Recovery 1 #Backup 2 #OpenAI经济研究 1 #职业边界 1 #跨岗位任务 1 #AI与工作 1 #Workforce Strategy 1 #OpenAI Economic Research 1 #Occupational Boundaries 1 #Task Crossover 1 #AI and Jobs 1 #LLM评测 7 #模型量化 1 #端侧AI 2 #LLM Benchmarks 1 #MLX 4 #iPhone AI 2 #Model Quantization 1 #On-device AI 1 #1-bit LLM 2 #PrismML 2 #Bonsai 27B 2 #版权 1 #演员肖像 1 #生产审批 1 #影视制作 1 #Netflix 1 #Copyright 2 #Talent Likeness 1 #Production Approval 1 #Content Production 1 #Netflix AI Guidelines 1 #Jensen-Shannon散度 1 #AI可观测性 1 #模型验证 1 #API漂移 1 #行为指纹 1 #LLM监控 1 #Jensen-Shannon Divergence 1 #AI Observability 1 #Model Verification 1 #API Drift 1 #Behavioral Fingerprints 1 #LLM Monitoring 1 #系统提示词 1 #Prompt工程 1 #System Prompts 1 #GPT-5.6 2 #LLM评估 3 #回归测试 4 #AI Agent测试 1 #CI/CD 1 #LLM Evals 4 #Regression Testing 4 #AI Agent Testing 1 #Claude Cookbook 2 #生产就绪 1 #软件测试 1 #Agent Swarm 1 #Production Readiness 1 #Software Testing 1 #SQLite 2 #Agent Swarms 1 #minisqlite 2 #无人机安全 1 #AI飞行控制 1 #Safety Case 2 #F-16 VENOM 2 #DARPA 2 #Project Pilot 2 #Drone-Bench 2 #Drone Safety 1 #AI Flight Control 1 #Prompt优化 1 #评估器 1 #LLM优化 1 #Prompt Optimization 1 #Evaluators 1 #LLM Optimization 1 #optimize_anything 2 #GEPA 2 #数据治理 3 #健康数据隐私 1 #医疗记录 1 #Data Governance 3 #EHR 2 #Apple Health 2 #Health Data Privacy 1 #Medical Records 1 #ChatGPT Health 2 #AI运营 1 #Agent评测 11 #人工升级 2 #Agent治理 7 #多 Agent 1 #证明验证 1 #数学证明 1 #推理优化 1 #基准测试 1 #CPU性能 1 #安全 1 #快照 1 #持久化 1 #状态管理 1 #Agent Evaluation 8 #Human Escalation 1 #AI Operations 1 #Agent Governance 6 #Enterprise AI Agents 1 #OpenAI Presence 2 #AI4Math 2 #Multi-Agent Systems 4 #Proof Verification 1 #Mathematical Proofs 1 #Inference Optimization 1 #tiktoken 2 #Hugging Face 4 #Benchmarking 2 #CPU Performance 1 #GigaToken 2 #Tokenization 2 #Durable Storage 1 #Security 1 #State Management 1 #Snapshots 1 #Persistence 1 #预训练模型 1 #推理性能 1 #多语言AI 1 #模型迁移 2 #开源AI 1 #AI产业 1 #中国AI 1 #开放权重 3 #Open Source AI 1 #Kimi 2 #DeepSeek 6 #Qwen 2 #China AI 1 #Open Weights 4 #Pretrained Models 1 #Inference Performance 1 #Multilingual AI 1 #Model Migration 2 #BPE 2 #Tokenizer 2 #GitHub Copilot 4 #Gemini CLI 2 #OpenAI Codex 2 #SKILL.md 4 #Agent Skills 5 #依赖安全 1 #沙箱 3 #包管理器 1 #安装前安全 1 #软件供应链 1 #AI Coding Agent 2 #Dependency Security 1 #Sandbox 4 #Package Managers 1 #Pre-Install Security 1 #Software Supply Chain 1 #AI Coding Agents 8 #银行生产力 1 #金融AI 1 #Banker Productivity 1 #Financial AI 1 #ChatGPT 4 #Singular Bank 2 #人机协作 2 #AI信心 1 #决策 1 #批判性思维 1 #自动化偏差 1 #AI建议 1 #Human AI Collaboration 1 #AI Confidence 1 #Decision Making 1 #Critical Thinking 1 #Automation Bias 1 #AI Advice 1 #软件架构 1 #可验证性 6 #Agent可靠性 4 #领域特定语言 1 #Software Architecture 1 #Verification 9 #DSL 2 #Agent Reliability 3 #Domain-Specific Language 1 #Model Context Protocol 2 #OAuth 2 #Human in the Loop 3 #MCP Elicitation 2 #长期记忆 1 #Agent可观测性 5 #Agent记忆 2 #因果评估 1 #教育AI 1 #学习成果 1 #教师AI 1 #Browser Agent 1 #Long-Term Memory 1 #Agent Observability 4 #AI Agent Memory 2 #Causal Measurement 1 #Claude for Teachers 2 #ChatGPT for Teachers 2 #Education AI 1 #Learning Outcomes 1 #AI for Teachers 1 #Web Bot Auth 4 #FP-Agent 2 #Cloudflare Precursor 2 #Bot Detection 2 #Browser Agents 1 #LLM安全 2 #LLM Security 2 #CVSS 2 #JEF 2 #CJS 2 #Jailbreak 2 #全局工作区 1 #Global Workspace 1 #LLM Interpretability 1 #Jacobian Lens 2 #J-space 4 #超级计算机 1 #Supercomputer 1 #以太网 2 #Ethernet 2 #RDMA 4 #OCP 2 #分布式训练 2 #Distributed Training 2 #网络协议 1 #Networking 1 #MRC 2 #前沿企业 1 #Frontier Firms 1 #AI素养 1 #组织转型 2 #AI技能缺口 1 #员工技能提升 1 #AI培训 1 #AI Literacy 1 #Organizational Transformation 2 #AI Skills Gap 1 #Workforce Upskilling 1 #AI Training 1 #OpenAI Academy 2 #On-Call自动化 1 #代码审查 3 #AI研究 1 #推理模型 1 #形式化验证 5 #单位距离问题 1 #离散几何 1 #AI数学 3 #AI Research 1 #Reasoning Models 1 #On-Call Automation 1 #Formal Verification 6 #Erdős 2 #Unit Distance Problem 1 #Code Review 5 #Discrete Geometry 1 #Ramp 2 #AI Math 1 #数据主权 1 #本地部署 1 #混合云 1 #AI Factory 2 #Data Sovereignty 1 #On-Premise AI 1 #Hybrid Cloud 1 #Dell 2 #身份验证 1 #钓鱼防护 1 #账户安全 2 #WebAuthn 2 #Authentication 1 #YubiKey 2 #Phishing 1 #FIDO 2 #Passkeys 2 #Account Security 2 #机器学习 1 #生产系统 1 #AI工程 1 #Agentic AI 6 #Machine Learning 1 #Production Systems 1 #AI Engineering 1 #Southeast Asia 1 #Elevation 1 #权限提升 1 #东南亚 1 #Sandbox Isolation 1 #沙箱隔离 2 #Shopee 2 #AI编码 2 #Windows Sandbox 1 #Windows 沙箱 1 #Codex 21 #Sea Limited 2 #生产力 1 #productivity 1 #Coase 2 #组织理论 1 #organizational theory 1 #工作未来 1 #货币政策 1 #monetary policy 1 #战略资源 1 #strategic resources 1 #token经济学 2 #地缘政治 1 #geopolitics 1 #AI转型 1 #财务规划 1 #企业财务 1 #AI Transformation 1 #Financial Planning 1 #CFO 2 #Enterprise Finance 1 #PwC 2 #AI规模化 2 #E2EE备份 1 #B2B Signals 4 #Labyrinth 2 #AI Scaling 2 #具身推理 1 #负责任扩展政策 1 #数据驻留 1 #机器人基础模型 1 #物理AI 1 #合规 1 #Responsible Scaling Policy 1 #Constitutional AI 2 #Embodied Reasoning 1 #Data Residency 1 #Motion Transfer 2 #SOC 2 2 #Robot Foundation Model 1 #Compliance 1 #Physical AI 1 #GPT-4 Enterprise 2 #Claude Enterprise 2 #劳动经济学 2 #AI采用 4 #观测暴露度 1 #AI劳动力市场 1 #Labor Economics 2 #监管框架 1 #AI 风险管理 1 #AI 合规 1 #AI 治理 1 #观察暴露度 1 #劳动力 1 #AI 采用 1 #AI 劳动力市场 1 #劳动力转型 1 #经济指数 3 #劳动力市场 2 #AI 研究计划 1 #经济未来 3 #Regulatory Framework 1 #Observed Exposure 1 #Workforce Transformation 1 #AI Risk Management 1 #Workforce 0 #AI Compliance 2 #Economic Index 2 #Labor Market 2 #ISO 42001 2 #AI Labor Market 1 #AI Research Program 1 #EU AI Act 3 #Economic Futures 2 #实时API 1 #全球规模 1 #Relay架构 1 #基础设施 1 #语音AI 3 #低延迟 1 #机械可解释性 1 #激活值口语化 1 #自然语言自编码器 1 #Global Scale 1 #Relay Architecture 1 #Pion 2 #Infrastructure 1 #NLA 2 #Realtime API 1 #Mechanistic Interpretability 1 #Low Latency 1 #Activation Verbalizer 1 #Kubernetes 2 #WebRTC 2 #Voice AI 3 #Natural Language Autoencoders 1 #稀疏自编码器 1 #多模态 1 #机器人 1 #音乐AI 1 #视频生成 1 #Multimodal AI 1 #AI Architecture 1 #Robotics 1 #Genie 1 #LLM Behavior 1 #Sparse Autoencoders 1 #Veo 1 #Gemini 4 #企业安全 2 #E2EE 备份 1 #硬件安全模块 1 #端到端加密 3 #Enterprise Security 2 #OPAQUE 4 #E2EE Backups 2 #Hardware Security Module 1 #HSM 4 #End-to-End Encryption 3 #WhatsApp 6 #垂直领域AI 1 #安进 1 #生物科技 1 #AI医疗 1 #药物发现 1 #生命科学 1 #Domain-Specific AI 1 #Moderna 2 #Amgen 1 #AlphaFold 2 #Biotech 1 #AI in Healthcare 1 #Drug Discovery 1 #Life Sciences 1 #GPT-Rosalind 2 #AGI 2 #数据中心 1 #Data Centers 1 #xAI 2 #Oracle 2 #人形机器人 1 #微软 1 #Nvidia 4 #英伟达 2 #英伟达GR00T 1 #GPU 2 #算力 1 #Compute 1 #具身智能 2 #AI基础设施 8 #AI Infrastructure 8 #AI机器人 2 #Stargate 2 #Humanoid Robots 1 #Atlas 2 #NVIDIA GR00T 3 #Physical Intelligence 4 #Boston Dynamics 2 #Embodied AI 2 #VLA 4 #AI Robotics 2 #Google DeepMind 7 #Gemini Robotics 4 #AI 政策 2 #国家安全 1 #可信访问 1 #AI 防御 1 #AI Policy 2 #National Security 1 #FedRAMP 2 #GPT-5.4-Cyber 2 #Trusted Access 1 #AI Defense 1 #开源权重 1 #数据合规 1 #PII 检测 1 #Agent 安全 1 #HIPAA 6 #GDPR 4 #AI 经济学 1 #自主交易 1 #Open Weight 1 #Data Compliance 1 #PII Detection 1 #Agent Safety 2 #Privacy Filter 2 #AI Economics 1 #Project Deal 2 #Autonomous Trading 1 #行为漂移 1 #奖励劫持 1 #强化学习 2 #RLHF 2 #Behavioral Drift 1 #Reinforcement Learning 2 #GPT-5 2 #AI 编程 2 #Claude Opus 4.7 2 #Frontier Math 2 #GPT-5.5 6 #开源安全 2 #零日漏洞 1 #网络安全 3 #Apple 2 #Microsoft 3 #Open Source Security 2 #Zero-Day Vulnerabilities 1 #Claude Mythos 2 #Cybersecurity 3 #Project Glasswing 2 #定性研究 1 #用户期望 1 #AI 安全 3 #多语言研究 1 #AI 采纳 1 #AI 用户研究 1 #Qualitative Research 1 #User Expectations 1 #Multilingual Study 1 #AI User Research 1 #护栏 1 #红队测试 2 #Guardrails 1 #Red Teaming 2 #Anthropic RSP 2 #OWASP LLM 2 #NIST AI RMF 4 #潜空间扩散 1 #生成式AI 2 #AI音乐 1 #Magenta 2 #Google 6 #SynthID 4 #Latent Diffusion 1 #Generative AI 3 #AI Music 2 #Lyria 3 #DeepMind 4 #视觉AI 1 #上下文工程 2 #AI原生工程 3 #提示词工程 1 #AI工作流 1 #AI治理 11 #AI战略 2 #企业AI 12 #认知技能 1 #AI采纳 2 #AI编程助手 1 #开发者生产力 5 #Context Engineering 4 #AI-Native Engineering 3 #Prompt Engineering 2 #AI Workflow 1 #Vision AI 1 #Cognitive Skills 1 #AI Governance 15 #AI Coding Assistants 1 #Opus 4.7 2 #AI Adoption 7 #Developer Productivity 6 #AI 战略 1 #企业 AI 2 #AI 产品 1 #API 2 #Claude Cowork 2 #Enterprise AI 15 #AI Products 1 #原型设计 1 #视觉协作 1 #AI 设计工具 1 #Figma AI 2 #Google Stitch 2 #Prototyping 1 #UI/UX 0 #Visual Collaboration 1 #AI Design Tools 1 #Claude Design 4 #Agent基础设施 2 #分布式推理 1 #边缘计算 1 #GPT-5.4 2 #Workers AI 2 #Distributed Inference 1 #Edge Computing 1 #Cloudflare 6 #AI编程 8 #SWE-bench 6 #LLM 14 #Benchmark 6 #AI Coding 11 #Sonnet 4.6 2 #渐进式发布 1 #站点可靠性 1 #金丝雀部署 1 #配置安全 1 #Progressive Rollout 1 #Site Reliability 1 #DevOps 2 #Canary Deployment 1 #Configuration Safety 1 #Meta 10 #AI Strategy 5 #Engineering Productivity 1 #LLM Cost 1 #Token Economics 3 #Context Infrastructure 2 #认知资产 1 #效率工程 1 #AI成本 1 #Token经济 1 #Sycophancy 1 #Reward Hacking 5 #Emotion Vectors 2 #Interpretability 4 #AI Safety 20 #情绪向量 2 #可解释性 5 #领导力 1 #组织重构 1 #AI原生 1 #企业转型 1 #组织设计 2 #Leadership 1 #Restructuring 1 #AI-Native 1 #Enterprise Transformation 1 #Organization Design 1 #自评 1 #技能 1 #职业替代 1 #未来工作 3 #职业 1 #Self-Assessment 1 #Skills 1 #Job Replacement 1 #Future of Work 5 #Career 1 #Agent 架构 1 #Agent Architecture 2 #OpenAI 64 #知识图谱 2 #数字化转型 2 #AI Agent 24 #企业软件 1 #Neo4j 2 #Knowledge Graphs 1 #Digital Transformation 2 #Enterprise Software 1 #SaaS 2 #Klarna 2 #Agent 基础设施 1 #互操作性 1 #钉钉 1 #飞书 1 #协议 1 #Agent Infrastructure 3 #Interoperability 1 #DingTalk 1 #Feishu 1 #Protocol 1 #AI Agents 25 #CLI 2 #MCP 14 #Software Engineering 4 #Constraints 1 #Claude Code 16 #Cursor 6 #软件工程 4 #约束系统 1 #工程化 1 #Harness Engineering 27 #安全架构 5 #运行时治理 2 #Agent安全 11 #AI安全 27 #Security Architecture 4 #Runtime Governance 1 #Agent Security 9 #AI Security 12 #学习曲线 1 #经济影响 1 #Learning Curves 1 #Economic Impact 1 #Claude 29 #Anthropic 54 #Programmers 1 #架构 2 #预防 2 #猝死 2 #健康 2 #认知 2 #程序员 1 #历史 2 #工程 3 #Agent 6 #AI 30 #Halo 0
调研 #SynthID#AI Watermarking#Content Provenance#Verification Throughput#Google DeepMind#C2PA#AI Safety

SynthID at Scale: Design the Watermark Verification Throughput Architecture

SynthID scale depends on verification throughput: version routing, uncertainty, audit evidence, review queues, and cost per defensible verdict.

Administrator Administrator Published on 2026-10-18
调研 #AI Search#Citation Security#RAG Security#GEO#Source Provenance#AI Safety

AI Search Citations Have a Supply-Chain Attack Surface

A study of 17,211 citations shows how source preference can amplify ordinary web publishing into an AI search security path.

Administrator Administrator Published on 2026-10-16
调研 #AI Safety#Third-Party Assessment#Independent Evaluation#AI Governance#Frontier Models

Third-Party AI Assessments Need Three Contracts

Third-party AI assessment needs enforceable contracts for independence, access, and disclosure, not a credibility label.

Administrator Administrator Published on 2026-10-16
解读 #AI Self-Replication#Autonomous Replication#AI Safety#IoT Security#AI Worms#Containment#Agent Evaluation

AI Self-Replication Without Datacenters: A Capability-to-Containment Evidence Contract

AI self-replication needs a five-layer evidence contract covering capability, resources, persistence, propagation, and containment.

Administrator Administrator Published on 2026-10-13
解读 #LLM Authority Bias#RAG Evaluation#Source Verification#AI Safety#Retrieval#Tool Outputs#NeurIPS 2026

Authority Bias in LLMs: When a Verified Source Defeats a Correct Answer

A verified-source label can overturn correct LLM answers. RAG systems need claim-level evidence and conflict tests, not source prestige alone.

Administrator Administrator Published on 2026-10-02
调研 #MentalHealthBench#Mental Health AI#AI Safety#LLM Evaluation#Clinical AI#Risk Metrics#OpenAI

MentalHealthBench: Build a Harm-Sensitive Evaluation Contract

Use MentalHealthBench as a diagnostic, then add severe-failure rates, escalation tradeoffs, worst-slice floors, independent review, and deployment gat

Administrator Administrator Published on 2026-09-24
解读 #AI Kill Switch#AI Governance#AI Incident Response#Frontier AI#Independent Verification#AI Safety#Recovery Protocol

AI Kill Switch Governance: Authority, Evidence, Recovery

An AI kill switch needs named authority, graded evidence, scoped containment, preserved telemetry, and a tested recovery path.

Administrator Administrator Published on 2026-09-20
解读 #Model Misalignment#AI Incident Reporting#Reproducible Evidence#AI Safety#OpenAI#Anthropic#Independent Evaluation

Model Misalignment Reporting Needs a Reproducible Evidence Contract

Model misalignment reporting needs replayable evidence, calibrated claims, independent access, and explicit closure criteria.

Administrator Administrator Published on 2026-09-17
解读 #EvoUndo#Self-Modifying Agents#Agent Harness#Recoverability#Rollback#AI Safety#LLM Agents

EvoUndo: A Recoverability Gate for Self-Modifying AI Agent Harnesses

EvoUndo shows why capability gains are insufficient for self-modifying Agent Harnesses. Require witnessed, counterfactual recovery before merge.

Administrator Administrator Published on 2026-09-02
解读 #Claude Fable 5.1#Claude Mythos 5.1#AI Safety#LLM Benchmarking#Terminal-Bench#Model Evaluation#AI Governance

Claude Fable 5.1 vs Mythos 5.1: How to Benchmark the Safeguard Tax

Fable 5.1 and Mythos 5.1 share model weights but use different safeguards. Measure capability, intervention, cost, and access together.

Administrator Administrator Published on 2026-09-02
Previous Next