全部 AI 动态

holy fuck. so auto compact basically loads last 1M tokens into context by default on claude? and here i thought they were summarising and reducing the context loads. no wonder people used to burn anywhere from 2%-15% 5h limits simply for auto compacting their sessions.

We just let in every Pro and Max user from the Claude Code Projects waitlist! If you're new to Claude Code Projects, here's a 4 minute walkthrough to get you started:

Grok Bot now has its own email. Bot can use it to sign up for services, contact businesses for you, or schedule time with someone.

@StepFun_ai just announced Step 5 Preview, and it's FREE in Kilo for a week: • 600B parameters, 27B active per token • 1M-token context + vision • highlights: strong coding and finance performance at lower cost. Try it and let us know what you think!

Please set up your email, @bot

Grok Bot now has its own email. Bot can use it to sign up for services, contact businesses for you, or schedule time with someone.

Grok Bot now has its own email. Bot can use it to sign up for services, contact businesses for you, or schedule time with someone.

Had @Muse call every Audi dealer near me to find a Q9 to test drive. It talked to real people and reported back. No Q9 on lots yet, but one guy invited us to a reveal event. Skipped an afternoon of hold music entirely. @alexandr_wang @AIatMeta @Audi

An update for builders using Codex on Windows: We’ve built a new sandbox mode using @Microsoft’s Execution Containers (MXC) for faster setup, stronger network enforcement, and granular file access controls. Requires a compatible Windows 11 device.

New ARC Prize 2026 - ARC-AGI-2 High Score 88.06% by @tufalabs The $150K Bonus Prize, on top of the guaranteed prizes, will be split among all teams scoring over 85%

Today, we announced new @Windows platform capabilities and @surface devices built to empower people and organizations for the era of hybrid intelligence. Here are some of the highlights: - It all starts with trust. Microsoft Execution Containers (MXC) is now generally available on Windows 11, keeping agents contained within boundaries the operating system enforces. - HydraFusion brings hybrid intelligence to life, allowing models to run locally while seamlessly accessing the cloud when more scale or capability is needed. We're working with @github to extend HydraFusion to Windows. - This fall, Search on the taskbar adds thousands of quick actions, making it easier to adjust settings and get things done as fast as you can type. - @MSFTCopilot is getting more powerful on Windows. With hybrid intelligence, it can tap into your PC's local capabilities and turn your requests into action, while keeping you in control. - Local agents are coming to Windows too. Muse is on its way as a native Windows app with MXC integration, and Hermes, OpenClaw, and Perplexity are already building local experiences on Windows. - Surface Laptop Ultra, built with NVIDIA RTX Spark, is the most powerful Surface we've ever built. It's available for pre-order today, along with the Surface RTX Spark Dev Box and new RTX Spark PCs from our partners. Call of Duty is coming to RTX Spark in 2027. Read more on the Windows Blog: https://blogs.windows.com/windowsexperience/2026/10/07/building-windows-for-hybrid-intelligence/

You can now train your own Decision model like Jev locally! We increased Qwen3.5 0.8B’s aggregate accuracy from 20.7% to 74.3% across 3 decision benchmarks - on just 4GB VRAM. Turn any LLM like Qwen3.8, Gemma 4 into decision models with our open-source Unsloth repo. We fine-tuned with a Clef head using Unsloth and LoRA (r=64) for one epoch, increasing downstream accuracy from 30–37% to 78%. GitHub: https://github.com/unslothai/unsloth Guide and Notebooks: https://unsloth.ai/docs/basics/train-your-own-decision-model-with-unsloth

we've raised $50M from Khosla Ventures, Accel, Initialized, Kevin Weil, DST Global to build the world's most wearable BCI cap at Sabi read more here: https://www.forbes.com/sites/rashishrivastava/2026/10/09/vinod-khosla-backs-an-ai-startup-building-a-mind-reading-baseball-cap/

Computer created Alexandria, a source-backed encyclopedia with 66,574 entries. It took 2.5 million credits, as Computer scanned and synthesized 348,980 sources. Alexandria is a free website, open to all, and will be maintained by Computer: https://alexandria.pplx.app

Meet Qwen-Image-2.1, the most balanced and cost-effective image generation model in the Qwen-Image series! Now open weights! 🎨 A unified model for both generation and editing, delivering top-tier quality in a lightweight package. Highlights: 👀 - Compact & exceptionally fast: A lightweight 7B architecture that outperforms most closed-source models, with drastically accelerated inference for multi-image inputs. - Native transparency: Natively generates and edits RGBA layers, enabling seamless compositing and text editing within transparent images. - Versatile, high-fidelity editing: Supports up to 10 reference images and precise local control while preserving strict fidelity for portraits and products. - Broad coverage & stunning aesthetics: Excels at panoramas, infographics, and virtual try-ons, delivering realistic textures and elegant typography. Start to create your next masterpiece with Qwen-Image-2.1! 🖼️ - Blog: https://qwen.ai/blog?id=qwen-image-2.1 - GitHub: https://github.com/QwenLM/Qwen-Image-2.1 - Model Scope: https://www.modelscope.cn/models/Qwen/Qwen-Image-2.1 - Hugging Face: https://huggingface.co/Qwen/Qwen-Image-2.1

A note from our research leaders: Last week we parted ways with Jasmine, Mikita, and Tomek after a thorough investigation found they violated clear policies on handling sensitive information. Our internal investigation uncovered a significant breach of trust beyond what’s outlined in the letter they published and we stand by the decision to not continue their employment. We generally keep individual employment matters private and don't believe a back and forth would be productive or lead to a resolution, but we want to address the points they raised in their letter directly. - We want to be very clear that these decisions were not about raising safety concerns or speaking out. Safety and research debates happen every day at OpenAI, often spirited and highly critical. We actively encourage these discussions and consider them essential to making the right decisions. We cannot do the work in front of us without a high degree of trust. We will continue to be extremely forgiving of our team making good-faith mistakes. We have not and do not terminate any of our employees for raising concerns. - We are actively finalizing contracts with third-party safety assessors and will announce details in the coming weeks. People across the company have been working really hard on getting these partnerships up and running. We are committed to embedding external assessors and continue to make close collaboration with independent safety organizations a core part of our safety work. Many of our researchers already work with 3p safety organizations productively. - We agree with the letter that preserving the monitorability of frontier models requires an industry-wide commitment, including from OpenAI. Monitorability has long been a core piece of our research program, and something we continue to invest significant resources in (see our publications on Monitoring Monitorability and the subsequent open sourcing of monitorability evals, our system card for GPT-6 Astra, Jakub’s blog and post on X, and the numerous blog posts on our Alignment blog on the topic). We are deeply sad about this outcome. We appreciated Jasmine, Mikita, and Tomek’s contributions to AI safety at OpenAI and their willingness to speak up and challenge ideas. We championed their voices, supported their work, and placed enormous trust in them. These decisions were not about them raising safety concerns. We have always encouraged that and always will.

Running MiniCPM5 from @OpenBMB on iPhone Air at 37 tok/s. One of the best open models at 2B parameters, now available in the NobodyWho Chat app.