🤖 AI 资讯日报 — 2026-07-10

数据来源: 12 个数据源 | 原始 78 条 → 去重 78 条 → 精选 63 条

📡 T1 官方一手

Anthropic Research

• A global workspace in language models
• Interpretability research on Claude's internal thoughts.

• Research
• Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.

• When AI builds itself
• For most of AI's history, humans drove every step in its development cycle. But at Anthropic, we are delegating a growing share of AI development to AI

• Home \ Anthropic
• Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.

GitHub Blog

• Automating cross-repo documentation with GitHub Agentic Workflows
• How the Aspire team turns merged product changes into SME-reviewed docs pull requests, closing the gap between release and documentation.

• Evaluating performance and efficiency of the GitHub Copilot agentic harness across models and tasks
• Explore how the GitHub Copilot agentic harness delivers strong results across multiple benchmarks and leading token efficiency.

Hugging Face Blog

• Trending Papers - Hugging Face
• ·Published on Jul 6, 2026. !

• State of Open Source on Hugging Face: Spring 2026
• # [](https://huggingface.co/blog/huggingface/state-of-os-hf-spring-2026#state-of-open-source-on-hugging-face-spring-2026) State of Open Source on Hugging Face: Spring 2026. [Established American compa

NVIDIA AI Blog

• NVIDIA Unlocks AI Compute at Scale, Inviting Partners to Power the AI Infrastructure Buildout
• NVIDIA is partnering with AI clouds to deploy large‑scale, multi‑tenant AI factories, aligning economics through a revenue-sharing and

• How Open Models Are Driving AI Research
• NVIDIA open models from Nemotron, Cosmos and BioNeMo are fueling the field's biggest research questions at ICML 2026.

• NVIDIA GTC 2026: Live Updates on What’s Next in AI
• Rolling coverage from San Jose, including NVIDIA CEO Jensen Huang's keynote, news highlights, live demos and on‑the‑ground color through

• What Are AI Tokens? The Language and Currency Powering Modern AI
• Tokens are units of data processed by AI models during training and inference, enabling prediction, generation and reasoning.

• AI Innovators Adopt NVIDIA Vera — Why Max Single-Threaded CPU at Scale Matters
• NVIDIA Vera exemplifies a new class of CPU, architected for the era of agents and being adopted by AI innovators including Perplexity;

OpenAI Blog RSS

• GPT-5.6 is now the preferred model in Microsoft 365 Copilot

• ChatGPT is now a partner for your most ambitious work

• GPT-5.5 Bio Bug Bounty

• GPT-5.6: Frontier intelligence that scales with your ambition

• Our approach to government and national security partnerships

• Separating signal from noise in coding evaluations

• Helping K–12 educators build practical AI skills

• Introducing GPT-Live

• MUFG aims to become AI-native with OpenAI

• Australian Payments Plus moves faster with ChatGPT and Codex

• How ChatGPT adoption has expanded

• Introducing GeneBench-Pro

• Core dump epidemiology: fixing an 18-year-old bug

• Inside Genebench-Pro

• Mapping Europe’s AI Workforce Opportunity

• HP Inc. launches Frontier strategic partnership with OpenAI

• Previewing GPT-5.6 Sol: a next-generation model

• How agents are transforming work

• OpenAI and Broadcom unveil LLM-optimized inference chip

• Helping build shared standards for advanced AI

Simon Willison

• Meta’s new model is Muse Spark, and meta.ai chat has some interesting tools
• Meta's new model is Muse Spark, and meta.ai chat has some interesting tools. 8th April 2026. Meta announced Muse Spark today,

• The new GPT-5.6 family: Luna, Terra, Sol
• 9th July 2026. OpenAI's latest flagship model hit general availability this morning, and comes in three sizes: Luna, Terra, and Sol (from

• Introducing Showboat and Rodney, so agents can demo what they’ve built
• Introducing Showboat and Rodney, so agents can demo what they've built. 10th February 2026. A key challenge working with coding agents is

xAI

• SpaceXAI
• Products. Solutions. Developer. Company. Pricing. [News](http

• Announcing xAI for Government | SpaceXAI
• We are excited to announce xAI For Government – a suite of frontier AI products available first to United States Government customers. Today, we’re proud to announce xAI For Government – a suite o

• xAI Raises $20B Series E | SpaceXAI
• Strategic investors in the round include NVIDIA and Cisco Investments, who continue to support xAI in rapidly scaling our compute infrastructure and buildout of the largest GPU clusters in the world.

📡 T1.5 媒体精选

HackerNews

• XGBoost beat LLMs at finding civilian-harm posts in Ukraine war Telegram data
• 👍 6 points

ITHOME

• NAIRR Science Program Reshapes Scientific Research, Powered by NVIDIA AI Infrastructure
• 美国国家科学基金会的NAIRR 试点项目在NVIDIA AI 基础设施支持下，推动了多个领域超过700 个创新科研项目，本文介绍了部分代表性项目。

• NVIDIA Vera Rubin Delivers World-Class Supercomputers for Science
• IT之家 6 月22 日消息，. NVIDIA Vera Rubin Delivers World-Class Supercomputers for Science. With 7 Exaflops of AI for Science and 5 Petaflops

• NVIDIA Announces BioNeMo Agent Toolkit — Tools for Agents to Accelerate Scientific Discovery
• IT之家 6 月23 日消息，. NVIDIA Announces BioNeMo Agent Toolkit — Tools for Agents to Accelerate Scientific Discovery. News Summary:.

THE DECODER

• Copilot goes cheap as Microsoft phases out OpenAI and Anthropic models to cut costs
• Microsoft is replacing AI models from OpenAI and Anthropic with its own MAI models in products like Excel and Outlook.

• Nvidia's Kyber NVL144 reportedly pushed back more than a year, Asian suppliers drop
• Nvidia's next AI server rack, Kyber NVL144, has been delayed more than a year to 2028 because of circuit board manufacturing problems,

• ChatGPT can now listen and talk at the same time, making AI conversations seem more human
• OpenAI's GPT-Live can listen and speak at the same time using a full-duplex architecture. Complex questions get handed off to GPT-5.5 in the

• OpenAI's GPT-5.6 launches Thursday after a delay forced by the U.S. government
• OpenAI is launching GPT-5.6 on Thursday after the U.S. government lifted its release ban following additional testing.

• Anthropic's Claude Cowork AI agent is now available on mobile and web
• Anthropic is rolling out its AI agent Claude Cowork to mobile and web. Until now, the feature was limited to the desktop app.

📡 T2 社区动态 & 论文

ArXiv

• LLMs for Explainable AI: A Comprehensive Survey
• # LLMs for Explainable AI: A Comprehensive Survey. Large Language Models (LLMs) offer a promising approach to enhancing Explainable AI (XAI) by transforming complex machine learning outputs into easy-

• Large Language Models for Constructing and Optimizing Machine Learning Workflows: A Survey
• # Large Language Models for Constructing and Optimizing Machine Learning Workflows: A Survey. Building effective machine learning (ML) workflows to address complex tasks is a primary focus of the Auto

• Bridging Traditional Machine Learning and Large Language Models: A Two-Part Course Design for Modern AI Education
• # Bridging Traditional Machine Learning and Large Language Models: A Two-Part Course Design for Modern AI Education. This paper presents an innovative pedagogical approach for teaching artificial inte

• Modular Machine Learning: An Indispensable Path towards New-Generation Large Language Models
• Large language models (LLMs) have dramatically advanced machine learning research including natural language processing, computer vision, data mining, etc., yet they still exhibit critical limitations

• AI Learning Algorithms: Deep Learning, Hybrid Models, and Large-Scale Model Integration
• # AI Learning Algorithms: Deep Learning, Hybrid Models, and Large-Scale Model Integration. In this paper, we discuss learning algorithms and their importance across various applications, emphasizing t

HF Daily Papers

• RoboTALES: Learning Reasoning-Guided Robot Policies via Task-Aligned Simulated Futures
• Pretrained video generative models are promising backbones for visuomotor control, but their imagined futures often drift from task intent and are not reliably action-conditional. As a result, these m
• ❤️ 3 upvotes

• TESSERA v2: Scaling Pixel-wise Earth Foundation Models
• Pixel-wise Earth-observation (EO) foundation models are now achieving state-of-the-art performance via generated spatial embeddings. However, how these models scale and how best to spend a pretraining
• ❤️ 1 upvotes

• OmniTacTune: Policy-Agnostic Real-World RL for Tactile Residual Adaptation of Visual Policies
• Visual policies learned from human videos, teleoperation, and robot demonstrations offer scalable motion priors, but often fail in contact-rich manipulation, where success significantly depends on loc
• ❤️ 4 upvotes

• Token-Based Dual-view Fusion and Adaptation of Large Vision Models for Breast Cancer Classification
• Accurate breast cancer classification from mammography requires effective integration of complementary information from craniocaudal (CC) and mediolateral oblique (MLO) views, which provide a more com
• ❤️ 1 upvotes

• AgentLens: Production-Assessed Trajectory Reviews for Coding Agent Evaluation
• We present AgentLens, a production-assessed benchmark for interactive code agents. Most code-agent benchmarks reduce a run to a single bit -- did the task pass? -- but the people who actually use thes
• ❤️ 4 upvotes

• Wake up for Touch! Mask-isolated Tactile Alignment Learning in MLLMs
• Touch supplies the physical grounding needed to perceive intrinsic material properties, such as friction and compliance, that vision alone often cannot resolve. Recent efforts for equipping multimodal
• ❤️ 1 upvotes

• Single-Rollout Asynchronous Optimization for Agentic Reinforcement Learning
• Reinforcement learning (RL) is becoming increasingly important for post-training large language models (LLMs). Previous RL pipelines for LLMs were mostly synchronous and batch-interleaved, which is in
• ❤️ 6 upvotes

• Sparse Delta Memory: Scaling the State of Linear RNNs through Sparsity
• Linear attention models allow a fixed state size and a fixed amount of compute per token. However, due to their limited state size, linear attention models fall behind in long-context recall compared
• ❤️ 5 upvotes

• Imagined Rollouts are Kinematic, Not Dynamic: A Diagnosis of Long-Horizon World-Model Failure
• Long-horizon failure in world models is conventionally attributed to compounding error, a generic framing that does not distinguish what kind of error compounds. We propose a kinematic-vs-dynamic refr
• ❤️ 1 upvotes

• Automating the Design of Embodied Agent Architectures
• Embodied agents are typically built as hand-designed compositions of perception, memory, planning, and action modules. This modularity exposes a large architectural design space, but current systems s
• ❤️ 5 upvotes