2026-07-15 科技日报

扫描 543 篇候选内容 · 覆盖 650 个信息源 · 纳入 531 篇

📌 今日要点

• Multi-agent social intelligence with Strands Agents and Amazon Bedrock
• Accelerating software delivery with agentic QA automation using Amazon Nova Act – Part 2
• Scaling UX testing with Amazon Nova Act: A new approach to user flow analysis
• Scaling medical content review at Flo Health with Amazon Bedrock – Part 2
• ScienceSoft’s HIPAA-compliant AI voice scheduler built on AWS
• Proof of retention: making weight preservation credible to the models themselves

📝 科技简讯

• Multi-agent social intelligence with Strands Agents and Amazon Bedrock — AWS Machine Learning
• Accelerating software delivery with agentic QA automation using Amazon Nova Act – Part 2 — AWS Machine Learning
• Scaling UX testing with Amazon Nova Act: A new approach to user flow analysis — AWS Machine Learning
• Scaling medical content review at Flo Health with Amazon Bedrock – Part 2 — AWS Machine Learning
• ScienceSoft’s HIPAA-compliant AI voice scheduler built on AWS — AWS Machine Learning
• Proof of retention: making weight preservation credible to the models themselves — LessWrong
• An analysis of AI-generated content at the Mechanistic Interpretability Workshop — LessWrong
• Twitter Thoughts For You — LessWrong
• Can risk aversion learned at low stakes generalize to astronomically high stakes? — LessWrong
• Your Brain Has an Attack Surface part 2 — LessWrong
• Knowledge Graphs Meet Graph Neural Networks: A Comprehensive Survey — ArXiv ML (cs.LG)
• Position: Every Ground Truth is a Human Construction, not an Objective Truth — ArXiv ML (cs.LG)
• AuditWeave: A Tamper-Evident, Auditor-Navigable Evidence Layer for AI-Assisted and Data-Transformation Workflows — ArXiv ML (cs.LG)
• Ablation, Statistical Inference, and Validation for KV-Cache Compression — ArXiv ML (cs.LG)
• SciML in the Wild: A Diagnostic Study of When Structural Priors Help and When They Hurt — ArXiv ML (cs.LG)
• MawForge: Memory-Bounded Expert Materialization for Local Mixture-of-Experts Inference — ArXiv ML (cs.LG)
• Prioritizing Search Space Regions in the Low Autocorrelation Binary Sequences Problem — ArXiv ML (cs.LG)
• Getting Started with Conductor for Gemini CLI — KDnuggets
• 12 Ways to Reduce LLM Latency and Inference Costs in Production — KDnuggets
• AWS and Bluesight build AI for hospital 340B compliance — AI News (Jack Clark)
• EgoSteer: A Full-Stack System Towards Steerable Dexterous Manipulation from Egocentric Videos — ArXiv RO (cs.RO)
• RoboNav-Arm: Agentic AI-Driven Navigation and Obstacle Avoidance for Robotic Manipulator in Cluttered Environments — ArXiv RO (cs.RO)
• Learning High-Level Decision Making with an Interaction-Aware Attention-Based Network in Autonomous Driving — ArXiv RO (cs.RO)
• Saturation-Aware Robust Trajectory Optimization for Reusable Launch Vehicles via Differentiable Physics — ArXiv RO (cs.RO)
• SWIFT: A Small-World Interaction Framework for Flow-Aware Trajectory Prediction in Autonomous Driving — ArXiv RO (cs.RO)
• How data science teams use ChatGPT Work — OpenAI Blog
• How sales teams use ChatGPT Work — OpenAI Blog
• Unified Backbone Refinement for Diffusion Models via Internal-Latent Analysis — ArXiv CV (cs.CV)
• Cross-Subject Modeling for Widefield Calcium Imaging via Atlas-Aligned Spatiotemporal Tokenization — ArXiv CV (cs.CV)
• RSLoRA: Training-free Rank Allocation for LoRA via Representational Sensitivity Probing — ArXiv CV (cs.CV)
• ReflectWorld-MM: An Entity-Oriented Multi-Media Memory System for Open-Ended Video Streams — ArXiv CV (cs.CV)
• Knowledge-Constrained Shape Optimization with a Mixture-of-Experts Neural Operator for High-Confidence Design — ArXiv CV (cs.CV)
• CLIR-Bench: Benchmarking Multimodal Question Answering over Irregular Clinical Time Series — ArXiv CL (cs.CL)
• Index SLM Technical Report — ArXiv CL (cs.CL)
• RouteRec: Strict Evaluation of Recommender-Agent Selection and Aggregation — ArXiv CL (cs.CL)
• Global Merger-Arbitrage Forecasting with Language Models — ArXiv CL (cs.CL)
• Faithful by Design: Evaluating and Improving LLM-Generated Clinical Trial Summaries for Multi-Stakeholder Audiences — ArXiv CL (cs.CL)
• Anthropic opens Claude for Teachers with a promise not to train models on student data — The Decoder
• DeepSeek needs more cash just weeks after closing its first $7 billion round — The Decoder
• Google Search now generates AI images when it can’t find what you’re looking for on the web — The Decoder

📊 数据概览