https://ai-tokens.cn/news/news-meta-muse-20260922.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Meta 个人 AI 智能体 Muse 上线两周登顶 App Store，单日市值飙升近 2000 亿美元 https://ai-tokens.cn/news/news-gpt-6-astra-20260903.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 OpenAI 发布 GPT-6 Astra：「欢迎来到 AGI 时代」，API 定价 $10/$50 为上代 2.5 倍 https://ai-tokens.cn/news/news-gemini-38-flash-20260903.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Google 三周内再推 Gemini 3.8 Flash：介绍价 $0.75/$3.75，Cyber 网络安全变体齐发 https://ai-tokens.cn/news/news-workbuddy-open-platform-20260903.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 WorkBuddy 开放平台解析：用 Skill、Connector、专家与知识库给 AI 同事「配工具」 https://ai-tokens.cn/news/news-wechatpay-ai-card-20260831.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 微信支付 AI 专属卡上新：一张卡直连 DeepSeek Harness 与 OpenClaw https://ai-tokens.cn/news/news-anthropic-mhs-20260828.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Anthropic 发布 MHS 模型硬件标准：让 AI 像插 USB-C 一样操控真实世界设备 https://ai-tokens.cn/news/news-046cfcbc98c199af.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 射频计算：探索智能边界｜CNCC 2026 https://ai-tokens.cn/news/news-2b9227ae1c18986a.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Tenstorrent hits $6.2B valuation to scale open-weight AI chips https://ai-tokens.cn/news/news-d1fa3609444938bc.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 People are using AI to make their smart TVs dumb again https://ai-tokens.cn/news/news-ea3b81d8158f7d03.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Nvidia launches new open models to run agents in the enterprise and on PCs https://ai-tokens.cn/news/news-111c7ae55acd3d24.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 How Deepgram optimized voice AI at scale with AWS Trainium https://ai-tokens.cn/news/news-27312a2971983170.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Push-based KV transfer to improve TTFT and TPOT for disaggregated inference in vLLM https://ai-tokens.cn/news/news-d47baa3bafb6f81b.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Improve TTFT for multi-turn agentic workloads with bi-directional KV transfer in vLLM https://ai-tokens.cn/news/news-ca2be23ea0a1d609.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 NVIDIA Advances Agentic Security for Media, From European Public Service Broadcasting to Global Content Production https://ai-tokens.cn/news/news-e0ab2a2e4734ec53.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 MiniMax M3.1 Flash shows China's AI price race has no floor in sight https://ai-tokens.cn/news/news-bbe5566e8b52e5df.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Kling AI drops version 4.0 with stereo audio, multilingual lip sync, and a lighter Flash variant https://ai-tokens.cn/news/news-7bf27ea750e3c763.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 JB Pritzker thinks AI could help us live to 100 — or threaten us like nukes https://ai-tokens.cn/news/news-c161d6bab52698d1.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 A hacker broke into John Deere and SharkNinja with an army of AI agents https://ai-tokens.cn/news/news-c713b9c06b3a0249.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Democrats think cracking down on AI could give them an edge https://ai-tokens.cn/news/news-0bd5c2eef6999d8f.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 MiniMax Just Launched M3.1 Flash: 428B MoE Agentic Model at $0.15/M Input https://ai-tokens.cn/news/news-ce06fc7641056adc.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Stripe says AI agent purchases are adding up, but it won't share the numbers https://ai-tokens.cn/news/news-0abac2498d5e300a.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 The neocloud shakeout is coming, and Nvidia has the most to lose https://ai-tokens.cn/news/news-7ee89adc89387032.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 独家｜前Graphcore总裁卢涛加入铭芯启睿，要把LPU做成「高铁」 https://ai-tokens.cn/news/news-ca77df14e0ee6b1b.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 MemSilicon完成超 4 亿元 A 轮融资，宣布战略升级 https://ai-tokens.cn/news/news-67ac6606be7056c8.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Muse Wonder 建模大翻车，亚历山大·王辩护遭马斯克无情补刀 https://ai-tokens.cn/news/news-735fe766c314f04d.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 NVIDIA Open Sources Run:ai Under Apache 2.0 for Kubernetes GPU Orchestration https://ai-tokens.cn/news/news-da92b80397690dde.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Structured Authority Injection: Persistent Human Control for LLM Agents via Typed MCP Overrides https://ai-tokens.cn/news/news-0b0024271a7c2b59.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Rethinking User Encoders: Routing Diffusion States for Scenario-Adaptive User Representation https://ai-tokens.cn/news/news-d2da03dd040aaeef.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 From Events to Reasoning: Hierarchical Rule Composition over Explicit States https://ai-tokens.cn/news/news-9f9467f75484ebfa.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 ECHOES: Seeing Beyond the Current Image via Recursive History Compression and Privileged On-Policy Distillation https://ai-tokens.cn/news/news-4c5e025cfc8f17ca.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Sensitivity Analysis of Decision Rules in an Agent-Based Agro-Ecological Model https://ai-tokens.cn/news/news-434246a93fec1553.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 PRIME: Towards Verifiable Evidence-Grounded Reasoning in LLM Agents https://ai-tokens.cn/news/news-5f9cc4741ffa08d5.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 From Natural Language Intent to Executable Control: Hierarchical Feedback-Guided LLM Planning for Multi-component Cyber-Physical Systems https://ai-tokens.cn/news/news-456c7c8f7dcbfa06.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Null Controls Are Necessary but Not Sufficient: Evaluating Acceptance Guards for Scientific Discovery and Interpretability https://ai-tokens.cn/news/news-1cc7c278b01af977.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Learnable Splines for Unstructured Neural Operators https://ai-tokens.cn/news/news-fa74b0cc3895cd40.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Functional Welfare for AI Safety https://ai-tokens.cn/news/news-7cc78127c883714f.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Author-in-the-loop Agentic Reproducibility Auditing for a Machine Learning Conference https://ai-tokens.cn/news/news-f34ad45741b405ff.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Harness-Delta Attribution: Dissecting Gains From Harness Evolution https://ai-tokens.cn/news/news-8cad1d7071d35894.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Testing the Planning Claims of LLM-Based Agents against Formal Semantics https://ai-tokens.cn/news/news-2ffe5ada9085e1ad.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Are Large Reasoning Models Good Everyday Reasoners? A Comparison with Non-Reasoning Models on Logical Reasoning Tasks https://ai-tokens.cn/news/news-bf28c7aa2007aa01.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 FRACTAL: A Self-Similar Multi-Agent Orchestration Architecture with Formal Communication, Convergence, and Depth Bounds https://ai-tokens.cn/news/news-135cf67151de604c.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Embeddings Are Overpriced? A Frozen 16 MB Vocabulary for Small and Ternary Language Models https://ai-tokens.cn/news/news-7537d7e9b5fa1afe.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Where Prefill Expert Reduction Hurts MoE Models: A Few Prompt Tokens Carry Most of the Short-Answer F1 Loss https://ai-tokens.cn/news/news-5db63836ea0cc2dd.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 From Parallels to Divergence: Language Models' Learning of Action Nominal Constructions https://ai-tokens.cn/news/news-50ab1df727133df6.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 RealCLI: Toward a Unified Evaluation of Human-Agent Collaboration for CLI Agents https://ai-tokens.cn/news/news-22696dc86e61ff3d.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 AuditKIT: A Unified Evaluation Spine for Benchmarking, Judging, and Comparing Large Language Models and Agentic Systems https://ai-tokens.cn/news/news-a71c76ddaafd8e32.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 alkhder at ImageEval 2026 Shared Tasks: Constrained Selection and Rotation-Averaged Logit Scoring for Multimodal Hallucination Detection https://ai-tokens.cn/news/news-2693328898d9ec81.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Literal Lock-In: How Test Language Shifts Literal Attraction in Typed Decision Models on Bengali Idioms https://ai-tokens.cn/news/news-4ae3d57c2dbeb93e.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 Pearl-Chat: Engineering a native Luganda language model from first principles in JAX and Flax NNX https://ai-tokens.cn/news/news-1b333c653a1b4192.html AI导航 zh-CN 2026-10-20T02:59:05+08:00 HeRE: A Novel Benchmark for Hebrew Relation Extraction via LLM-Guided Denoising of Knowledge Graph Alignments