2026-08-18 科技日报

扫描 570 篇候选内容 · 覆盖 650 个信息源 · 纳入 555 篇

📌 今日要点

• How Much Memory Does Your Agent Actually Need?
• Multi-Vector (Late Interaction) Embedding Models with Sentence Transformers
• Amazon Bedrock AgentCore payments is now generally available: Enabling agents to transact safely and autonomously at scale
• Customize Amazon Quick embedded chat into your application
• Implement vector-prompt document classification using Amazon Bedrock
• How Jumio built a real-time feature store on AWS

📝 科技简讯

• How Much Memory Does Your Agent Actually Need? — Hugging Face Blog
• Multi-Vector (Late Interaction) Embedding Models with Sentence Transformers — Hugging Face Blog
• Amazon Bedrock AgentCore payments is now generally available: Enabling agents to transact safely and autonomously at scale — AWS Machine Learning
• Customize Amazon Quick embedded chat into your application — AWS Machine Learning
• Implement vector-prompt document classification using Amazon Bedrock — AWS Machine Learning
• How Jumio built a real-time feature store on AWS — AWS Machine Learning
• Improve contract search accuracy with auto-generated filters in Amazon Bedrock — AWS Machine Learning
• How Axonius built secure multi-tenant AI agents on Bedrock AgentCore — AWS Machine Learning
• Run Qwen3.8-27B as a Local AI Coding Agent in Just 3 Commands — KDnuggets
• 5 Things Vibe Coding Gets Right and 5 Things It Gets Wrong — KDnuggets
• 刚刚，GPT-6 代号再泄露！或将周四见 — 新智元
• 00 后高盛金融男，欲奸杀前女友，被 ChatGPT 上报后被捕 — 新智元
• 浙大开源 PhyEdit：单图编辑 + 精准 3D 物体操作 | ACM MM’26 — 新智元
• 冷战核废墟变 8 吉瓦 AI 超级工厂！黄仁勋押 15 亿，OpenAI 独霸 20 年 — 新智元
• GPT-5.6 Sol 暴涨 3 倍，OpenAI 最强视觉 AI 登顶！ — 新智元
• OpenAI president urges enterprises to hasten AI security defences — AI News (Jack Clark)
• Alvys launches AI agents for freight TMS workflows — AI News (Jack Clark)
• Reading Zhipu’s GLM-5.3 results past the headline number — AI News (Jack Clark)
• Learning Discrete Riemannian Metrics for Physical Fields with Cochain-Frame Equivarianc — ArXiv ML (cs.LG)
• Forward Pass Domain Adaptation (Without Cross-Layer Backpropagation) — ArXiv ML (cs.LG)
• Coarse-to-Fine Multi-Resolution Diffusion Models for Trajectory Generation in Urban Systems — ArXiv ML (cs.LG)
• Geometry Is Not Robustness: A Trajectory-Level Study of PGD Evaluation — ArXiv ML (cs.LG)
• DumpsterCluster: From Dumpster Diving to Serving LLaMA-70B on $60 GPUs — ArXiv ML (cs.LG)
• Calibrated Trust, Not Sharper Prediction: An Empirical Test of Uncertainty Fusion — ArXiv ML (cs.LG)
• PIKFNO: An Interpretable Neural Operator Based on Physics Informed Kernel Function — ArXiv ML (cs.LG)
• Cursor capitalizes on GitHub frustration, launches rival hosting platform — TechCrunch AI
• OpenAI institutes new safeguards after Hugging Face breach — TechCrunch AI
• Etched’s valuation doubles to $21B in a month — TechCrunch AI
• Why Apple’s camera-equipped AirPods may not be the ‘pervert pods’ consumers fear — TechCrunch AI
• Warp’s new system is an out-of-the-box software factory for AI development — TechCrunch AI
• OpenAI launches a safer ChatGPT for teens — years after teens started using it — TechCrunch AI
• Perplexity’s free AI offer left it with millions more users in India — TechCrunch AI
• Strengthening democratic oversight in national security — OpenAI Blog
• Partnering with CodeAI to prepare the first AI generation — OpenAI Blog
• Pacing model development in an era of cyber-critical capabilities — OpenAI Blog
• Introducing ChatGPT for Teens: Built for learning, backed by protections — OpenAI Blog
• How NVIDIA scales expertise with ChatGPT Work — OpenAI Blog
• Asana cleared 5 years of engineering work in 2 weeks with Codex — OpenAI Blog
• Anthropic Risk Report: August 2026 — LessWrong
• Funding Formal Methods for the Cyberpocalypse — LessWrong

📊 数据概览