2026-09-27 AI 日报

Multi-agent harness that runs Claude Code and Codex together as one system

Fast, exact LLM decoding on Apple Silicon (MLX) behind an OpenAI-compatible endpoint

The Office Harness for AI Agents — Spreadsheets, Docs, Slides, Canvas, Relational Tables, and PDF in one runtime.

Learn it. Build it. Ship it for others.

Turn any REST, SOAP, GraphQL, OData or SQL API into MCP tools for Claude & ChatGPT. Self-hosted. 265 connectors: SAP S/4HANA & Business One, ERP, e-commerce.

Model Context Protocol Server for Mobile Automation and Scraping (iOS, Android, Emulators, Simulators and Real Devices)

25 MB lightweight cross-platform database client for 100+ databases, including MySQL, PostgreSQL, SQLite, Redis, MongoDB, DuckDB, SQL Server, and Dameng.…

Source control for agents. Use multiple coding agents, track their changes and query them in one place

380 Claude Code skills & agent skills & plugins (30+ Agents, 70+ custom commands, 380+ skills, customizable references, scripts)for Claude Code, Codex, Gemini…

The open-source app everyone uses to manage agents at work

Unobserved common causes are pervasive in real-world time series and can induce spurious associations that causal discovery methods mistake for direct edges.…

Machine learning for scientific discovery is almost systematically data bound. Producing relevant high quality data, under budget constraints, is amongst the…

Tabular in-context learners such as TabPFN, Mitra, or ConTextTab rely on alternating row and column attention over 2D sequences of latent embeddings. These…

Continual spatio-temporal forecasting supports traffic management and environmental monitoring under evolving dynamics and expanding sensor networks. However,…

Learned particle simulators based on graph neural networks achieve strong one-step accuracy, but errors compound over long horizons. An underexplored variable…

Multi-teacher on-policy distillation (MOPD) integrates specialized capabilities into a single student, but existing practice typically hard-routes each prompt…

COLMAP-free 3D Gaussian Splatting (3DGS) bypasses computationally expensive structure-from-motion (SfM) pipelines, yet progressive camera pose tracking…

Difference-in-differences (DID) studies are widely used to evaluate climate policy, but assessing the evidence supporting their identification assumptions…

Building adaptable AI systems requires effective coordination of specialized capabilities across diverse tasks. Low-rank adaptation (LoRA) enables modular…

Ultrasound is one of the most widely used medical imaging modalities, and recent large vision-language models(VLMs) have shown increasing capabilities in…

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that…

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that…

Did you ever click on an “AI Arena” expecting glorious battle and instead get a boring benchmark? If so, this project is for you: proper life-or-death fights…

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that…

MiniMax-M3.1 Flash Preview is now live on the Token Plan! Faster, lighter, and built for teams running high-volume, latency-sensitive workloads, now available…

Meta came out with a banger paper https://arxiv.org/pdf/2606.00206, but it did not look at various quantizations supported in llama.cpp. So I did a run on 50…

AI Engineering from Scratch is an MIT-licensed curriculum: 523 lessons across 20 phases, from linear algebra and backprop to transformers, LLMs, agents, and…

I was reading a paper that surveyed the field of neural architecture search, where it said within 5 years, around 3000+ new models were proposed. The amount…

It had such a long run. It was my first introduction to modern language models. I remember getting slightly excited over it. And now it lives purely in our…

I run my own LLM while also having a Openai subscription. Also tried DeepSeek (latest flash now). I run Qwen 3.8 flash Next at an amazing speed on my 2x3090 +…

Six public decision benchmarks, one protocol: JEV-27B averages 84.07, TypeSafe Jev 1.13 83.85 Close to Jev at the level of whole probability distributions:…

Built using Claude Code as a local Chrome extension, it scans the text and rates it based on a number of questions. The box in the corner gives you a summary…

DeepSeek has 64% market share today 3 months ago, it had at 25% open models are winning https://t.co/ROsouLx1BZ

Trying something new - introducing CyberPVP - in other words CyberKimi vs other AI models competing to solve complex cyber tasks. you can see it live here:…

Meet GPT-6 Sol and Luna, two models that bring frontier intelligence to everyday work with different balances of capability and cost.

New Features Configure copy-on-select and right-click paste in the fullscreen TUI. Copied transcript selections now preserve Markdown formatting. (#47639,…

Learn how GPT-6 improves prompt caching with higher cache hit rates, new diagnostics, explicit breakpoints, and controls that reduce latency and costs.

Gemini 3.8 Flash-Lite TTS and Gemini 3.8 Flash TTS are our most expressive audio models yet.

Guest writer and physicist Matt von Hippel shares what happened when he issued a challenge to AI companies to solve a problem in his former subfield of…

We’re announcing a new life sciences research group and laboratory at Anthropic. This post introduces the team behind this work and shares early results in…

Google Cloud API Gateway now acts as a native remote Model Context Protocol (MCP) server, eliminating the need to build and maintain custom middleware to…

The Google Antigravity SDK now empowers developers to execute offline, agentic workflows locally using models like Gemma 4 26B A4B via LiteRT. This update…