🤖 AI 资讯日报 — 2026-06-18

自动生成于 Hermes Agent | 数据采集时间: 2026-06-18 08:32 UTC+8

📊 今日概览

🟢 T1 官方一手

🟢 OpenAI Blog

• Built to benefit everyone: our plan
• A vision for the future of AI, focusing on access, safety, and shared prosperity as OpenAI works to ensure AGI benefits everyone.

• Codex for every role, tool, and workflow
• Discover new Codex plugins, sites, and annotations that help analysts, marketers, designers, investors, and other teams get more done with

• Introducing GPT-5.5
• Introducing GPT-5.5, our smartest model yet—faster, more capable, and built for complex tasks like coding, research, and data analysis

🟣 Anthropic Blog

• 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

• Claude Fable 5 and Claude Mythos 5
• Today we're launching Claude Fable 5: a Mythos-class model that we've made safe for general use.

• Introducing Claude Opus 4.8
• Our latest model, Claude Opus 4.8, is an upgrade to our Opus class of models, with stronger performance across coding, agentic tasks,

• Claude Mythos 5
• Claude Mythos 5 is the latest update for Mythos Preview, seeing gains in cybersecurity, biology, and healthcare benchmarks. Given its capabilities, Mythos 5

• Agentic coding and persistent returns to expertise
• Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.

🤗 Hugging Face Blog

• GLM-5.2: Built for Long-Horizon Tasks
• A Blog post by Z.ai on Hugging Face.

• google/diffusiongemma-26B-A4B-it
• Hugging Face | GitHub | Launch Blog | Documentation License: Apache 2.0 | Authors: Google DeepMind. DiffusionGemma is a generative model

🐙 GitHub Blog

• What are git worktrees, and why should I use them?
• Git worktrees have been around since 2015, but it wasn't until recently they became popular. Learn what they are, how to use them, and why

• GitHub Code Quality generally available July 20, 2026
• More than 10,000 enterprises have used the GitHub Code Quality public preview to detect maintainability and reliability issues,

• GitHub Copilot CLI for Beginners: Overview of common slash commands
• GitHub Copilot CLI for Beginners: Learn how to use slash commands to control your terminal AI agent.

• Accelerating researchers and developers building multilingual AI with a new open dataset
• A new repository-level dataset, published on GitHub under CC0-1.0, helps researchers and developers discover multilingual developer content.

• GitHub Copilot app: The agent-native desktop experience
• At Microsoft Build 2026, GitHub introduced new tools, updates, and surfaces so agents can work the way you already work.

🍎 Apple ML

• Introducing the Third Generation of Apple's Foundation Models
• Our next generation of Apple Intelligence is centered around our users, integrated deeply into our operating systems, and powered by a bold…

• Work with us
• Build amazing machine learning experiences with Apple. Discover opportunities for developers and researchers at Apple.

• Instruction-Following Pruning for Large Language Models
• Our approach, termed "instruction-following pruning", introduces a sparse mask predictor that takes the user instruction as input and dynamically selects the

• IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2026
• Apple is presenting new research at the annual IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), which takes place in…

• Announcing the 2022 Apple Scholars in AIML
• The Apple Scholars in AIML PhD fellowship recognizes the contributions of emerging leaders in computer science and engineering at the graduate and postgraduate

💚 NVIDIA Blog

• Hands Free, AIs Forward: NVIDIA XR AI Brings Agents to AR Glasses
• NVIDIA XR AI is now available in public beta, giving developers a framework for building multimodal AI agents for AR glasses and XR devices.

• Coherent Breaks Ground on Expanded Texas Facility, Scaling AI's Optical Backbone
• The company makes the lasers, optical components and compound semiconductors that wire AI systems together — and runs what it calls the

• HPE AI Factory With NVIDIA Expands for the Era of Agents
• HPE AI Factory with NVIDIA expands with NVIDIA Vera CPU and NVIDIA Agent Toolkit for HPE Private Cloud AI — plus NVIDIA Confidential

• Fastest, Largest, Strongest: NVIDIA Blackwell Sweeps MLPerf Training 6.0
• Every breakthrough AI model starts the same way: with a training run. The infrastructure running those training jobs shapes everything: how

• Save Big and Play Bigger: GeForce NOW Summer Sale Brings Major Membership Savings
• Level up with the best value in gaming, 'Guild Wars' and rewards, plus eight new games this week.

✖️ xAI

• Grok Imagine Video 1.5 - xAI
• Grok Imagine Video 1.5 is now generally available on the Imagine API. Better motion, better physics, better audio, at the fastest speeds.

🧑‍💻 Simon Willison

• Anthropic Walks Back Policy That Could Have 'Sabotaged' AI Researchers Using Claude
• "We're changing Fable 5's safeguards for frontier LLM development to make them visible." Anthropic said in a statement to WIRED.

• Why AI hasn't replaced software engineers, and won't
• In March 2025, New York became the first U.S. state to add an AI disclosure checkbox to WARN Act filings. In the full first year,

• Uber Caps Usage of AI Tools Like Claude Code to Manage Costs
• The rideshare giant is limiting all employees to $1,500 in monthly token spending per AI coding tool, an Uber spokesperson said in response to a

• Your job is to deliver code you have proven to work
• 18th December 2025. In all of the debates about the value of AI-assistance in software development there's one depressing anecdote that I

🟡 T1.5 媒体 & 社区

🇩🇪 THE DECODER

• Landmark German ruling declares Google's AI Overviews are Google's own words and makes it liable for false answers
• A German regional court has ruled that Google is directly liable for the content of its AI search overviews. According to the court,

• Google Research's Gemini-SQL2 tops text-to-SQL benchmarks by a wide margin
• Google Research's Gemini-SQL2 turns natural language into executable SQL queries. Built on Gemini 3.1 Pro, it tops the BIRD benchmark at

• AI assistants fail basic fact-checking in BBC news study
• A systematic evaluation of leading AI chatbots reveals widespread problems with accuracy and reliability when handling news content.

• Algorithms from the 1980s power today's AI breakthroughs, earn Turing Award for researchers
• Andrew Barto and Richard Sutton have won the 2024 AM Turing Award for developing key technologies that power modern artificial intelligence.

🟧 HackerNews

• Ask HN: Is it time to fork HN into AI/LLM and "Everything else/other?" (553pts)
• I would very much like to enjoy HN the way I did years ago, as a place where I'd discover things that I never otherwise would have come across.

  The increasing AI/LLM domination of the site

• Coconut by Meta AI – Better LLM Reasoning with Chain of Continuous Thought? (362pts)

• Show HN: Countless.dev – A website to compare every AI model: LLMs, TTSs, STTs (361pts)

• Taste in the age of AI and LLMs (265pts)

• Wikimedia Enterprise – APIs for LLMs, AI Training, and More (222pts)

• Anti-AI Hype LLM Reading List (208pts)

• Jellyfin LLM/"AI" Development Policy (207pts)

• Ask HN: Go deep into AI/LLMs or just use them as tools? (195pts)
• I'm a software engineer with a solid full-stack background and web development. With all the noise around LLMs and AI, I’m undecided between two paths:

  1. Invest time in learning the internals

• Advent of Code 2023's new AI/LLM Policy (174pts)

• Should we use AI and LLMs for Christian apologetics? (2024) (170pts)

• RFC: Banning "AI"-backed (LLM/GPT/whatever) contributions to Gentoo (168pts)

• Benchmarks and comparison of LLM AI models and API hosting providers (152pts)
• Hi HN,

ArtificialAnalysis.ai provides objective benchmarks and analysis of LLM AI models and API hosting providers so you can compare which to use in your next (or current) project.

The site consol

• Pi.ai LLM Outperforms Palm/GPT3.5 (151pts)

• A guide to Gen AI / LLM vibecoding for expert programmers (128pts)

• Declarative Programming with AI/LLMs (121pts)

🔵 T2 学术 & 社交

🤗 Hugging Face Daily Papers

• LoopCoder-v2: Only Loop Once for Efficient Test-Time Computation Scaling ⬆️ 113
• Looped Transformers scale latent computation by repeatedly applying shared blocks, but sequential looping increases latency and KV-cache memory with the loop count. Parallel loop Transformers (PLT) al

• DreamX-World 1.0: A General-Purpose Interactive World Model ⬆️ 95
• DreamX-World 1.0 is a general-purpose interactive text/image-to-video world model for controllable long-horizon generation. It supports camera navigation, revisits to previously observed regions, and

• Ling and Ring 2.6 Technical Report: Efficient and Instant Agentic Intelligence at Trillion-Parameter Scale ⬆️ 69
• Efficient and scalable agentic intelligence requires models that can deliver both low-latency responses and strong reasoning capabilities while remaining practical to train, serve, and deploy. In this

• ACE-Ego-0: Unifying Egocentric Human and Robotic Data for VLA Pretraining ⬆️ 40
• Vision-Language-Action (VLA) models benefit from large-scale and diverse embodied data, yet scaling robot trajectory collection is costly and labor-intensive. Recent advances show that large-scale ego

• Zone of Proximal Policy Optimization: Teacher in Prompts, Not Gradients ⬆️ 40
• Knowledge distillation transfers a teacher's competence to a small student but is brittle in the small-student regime: forcing the student to imitate logits from a much larger teacher concentrates it

• GameCraft-Bench: Can Agents Build Playable Games End-to-End in a Real Game Engine? ⬆️ 36
• Game generation is an emerging application of coding agents, requiring models to transform natural-language specifications into playable interactive systems. Unlike traditional coding tasks, game gene

• LectūraAgents: A Multi-Agent Framework for Adaptive Personalized AI-Assisted Learning and Embodied Teaching ⬆️ 31
• Effective personalized AI-assisted learning demands systems that can not only generate accurate learner-specific educational materials, but also dynamically adapt their instruction to diverse learners

• Learning from the Self-future: On-policy Self-distillation for dLLMs ⬆️ 25
• On-policy self-distillation (OPSD) has proven effective for post-training large language models (LLMs), yet its application to diffusion LLMs (dLLMs) remains unexplored. Existing OPSD methods are inhe

• TRIAGE: Dialectical Reasoning for Explainable Risk Prediction on Irregularly Sampled Medical Time Series with LLMs ⬆️ 25
• Clinical early warning systems built on electronic health records, in which clinical observations are recorded as irregularly sampled medical time series (ISMTS), must deliver both calibrated risk sco

• OPD-Evolver: Cultivating Holistic Agent Evolver via On-Policy Distillation ⬆️ 21
• Memory has become a standard substrate for self-evolving agents, yet retaining experience is not the same as learning how to evolve through it. Existing memory agents can store trajectories, retrieve

📄 ArXiv 论文

• Artificial Intelligence
• Artificial Intelligence. Authors and titles for recent submissions. Wed, 17 Jun 2026 · Tue, 16 Jun 2026 · Mon, 15 Jun 2026 · Fri, 12 Jun 2026 · Thu,

• From AGI to ASI
• Tim Genewein Google DeepMind Matija Franklin Google DeepMind Alexander Lerchner Google DeepMind Laurent Orseau Google DeepMind Samuel

• A Survey of Large Language Models
• In this survey, we review the recent advances of LLMs by introducing the background, key findings, and mainstream techniques.

• The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery
• One of the grand challenges of artificial general intelligence is developing agents capable of conducting scientific research and

• Retrieval-Augmented Generation for Large Language Models: A Survey
• This comprehensive review paper offers a detailed examination of the progression of RAG paradigms, encompassing the Naive RAG, the Advanced RAG, and the

• Attention Is All You Need
• We propose a new simple network architecture, the Transformer, based solely on attention mechanisms, dispensing with recurrence and convolutions entirely.

• Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task
• This study explores the neural and behavioral consequences of LLM-assisted essay writing. Participants were divided into three groups: LLM, Search Engine, and

• Language Models are Few-Shot Learners
• Recent work has demonstrated substantial gains on many NLP tasks and benchmarks by pre-training on a large corpus of text followed

• Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks
• We explore a general-purpose fine-tuning recipe for retrieval-augmented generation (RAG) -- models which combine pre-trained parametric and non-parametric

• arXiv.org e-Print archive
• arXiv is a free distribution service and an open-access archive for nearly 2.4 million scholarly articles in the fields of physics, mathematics, computer

🐦 KOL 观点

• AI NEWS: A new open-source LLM that beats Grok, LLama-2, and Mixtral
• AI NEWS: A new open-source LLM that beats Grok, LLama-2, and Mixtral is here. Plus, more developments from Anthropic Claude 3, Amazon, MIT,

• this is extremely bad news for the AI industry
• this is extremely bad news for the AI industry... as we all know LLM development has been stagnant for many years and chatgpt will never get any

• AI Forward Deployed Engineer
• Right now, I see surging demand for AI Engineers who can build software applications using AI software components (like LLM prompting, agentic

• AI #169: New Knowledge | Zvi Mowshowitz
• Overcoming Bias. Censorship regimes by default get favorable LLM treatment. Get My Agent On The Line. Use your agent as a filter or commitment

• BREAKING: A new Claude Mythos 5 model slug spotted
• BREAKING: A new Claude Mythos 5 model slug has been spotted via Dev Mode. Claude Mythos is planned to be released as its own model class, besides Haiku,

• Claude Code is not AGI, but it is the single biggest advance in AI since the LLM
• Claude Code is not AGI, but it is the single biggest advance in AI since the LLM. But the thing is, Claude Code is NOT a pure LLM.

本报告由 Hermes Agent 自动生成 | 2026-06-18