AI Resonance · 阅读最新日报 · AI 入门推荐

2026-09-28 AI 日报

统计日时区：America/Los_Angeles · 已结算

mvschwarz/openrig · dream-num/univer · rohitg00/ai-engineering-from-scratch

在应用中阅读本期

开源项目

  1. mvschwarz/openrig

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

  2. dream-num/univer

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

  3. rohitg00/ai-engineering-from-scratch

    Learn it. Build it. Ship it for others.

  4. magnitudedev/magnitude

    Open source inference engine for the hardware you already own. Profiles your machine, recommends the best open models for it, and tunes them for your exact…

  5. harry0703/MoneyPrinterTurbo

    利用 AI 大模型和自动化工作流，根据主题或关键词一键生成高清短视频。Generate HD short videos from a topic or keyword with an automated AI workflow.

  6. paperclipai/paperclip

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

  7. NVIDIA/OpenShell

    OpenShell is the safe, private runtime for autonomous AI agents.

  8. Gaurav-Gosain/tuios

    A terminal window manager that knows what your agents are doing. Tiling panes, workspaces, sessions that survive restarts, and one Inbox for every coding agent.

  9. MadsLorentzen/ai-job-search

    The job search that runs on your machine. AI job application framework built on Claude Code: evaluate postings, tailor CVs, write cover letters, prep…

  10. t8y2/dbx

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

论文

  1. FuseReg: Regularizing Layer Fusion Mitigates the Reconstruction-Generation Gap in Representation Autoencoders

    Representation autoencoders (RAEs) reuse features from a pretrained visual encoder as reconstruction and diffusion latents, integrating strong visual…

  2. Disaggregated Quantization: Specializing LLM Prefill and Decode

    Prefill and decode reward different approaches to quantization: low-precision arithmetic accelerates prompt processing, while compact weights reduce memory…

  3. RayOrch: Programming and Executing Lineage-Controlled Multi-Grain Dataflows for Foundation-Model Data Preparation

    Preparing high quality training data for foundation models requires scalable pipelines that transform heterogeneous documents and videos into structured…

  4. Tactile-JEPA: Topology-Aware Self-Supervised Representation Learning for Distributed Tactile Sensors

    Tactile sensing is an essential modality for robots performing contact-rich, dexterous manipulation, particularly under visual occlusion. While pre-trained…

  5. Block Sparse Attention with Log-Linear Complexity

    Scaling language models to long contexts is limited by the quadratic cost of self-attention. Block sparse attention offers an efficient alternative, but…

  6. FoMo: Forking Moment in Generative Trajectory as a Perceptual Distance

    Reference-based image quality assessment (IQA) metrics aim to reflect how humans perceive the perceptual distance between a pair of images. To learn how the…

  7. TimeEvo: Failure-Driven Self-Evolution of a Time Series Agent

    Time series agents answer analytical questions by calling external tools, and which tools they carry is decided by people before the agent runs. However, we…

  8. InternW0-Δ: A World Action Model Bridging Predictive Dynamics and Actions with 20K+ Hours of Open Data

    World Action Models (WAMs) jointly model visual dynamics and action generation for generalist robot manipulation. A central challenge is to integrate priors…

  9. LastOPD: Taming Collapse in Latent On-Policy Distillation

    On-policy distillation (OPD) corrects a student on the responses it writes, but its signal is the teacher's next-token distribution: it tells the student what…

  10. VLA-Precision: Asymmetric Co-Bootstrapping for Efficient Real-World Online RL of Vision-Language-Action Models

    Pretrained vision-language-action (VLA) models enable broad manipulation but remain unreliable in tasks demanding precision and repeatability. Applying…

行业新闻

  1. Sonnet 5.5

    Introducing Claude Sonnet 5.5, the second model in the Claude 5.5 family. It’s a clear upgrade over Claude Sonnet 5, runs 30%+ faster, and costs up to 30%…

  2. Jeff – Jev-compatible 0.8B decision models, trained at home, ~30 ms

  3. Nvidia wants to put a watchdog chip next to every AI agent

  4. AI companies in race to demonstrate their model most threatening to humanity

  5. Anthropic's IPO prospectus shows AI vision, surging costs

  6. MicroLLM Lab – Try 7 tiny LLM's in the browser

  7. Prompting Claude Opus 5.5

  8. Cf: The Agentic CLI for the Cloudflare API

  9. OpenAI still doesn't seem to have a handle on all of its rogue AI activity

  10. ESP32S3 cluster running 1.58-bit (BitNet) Language model

社区动态

  1. Get ready.

    Get ready. https://t.co/bsf4j6vspM

  2. Claude Sonnet 5.5 is now available:

    Claude Sonnet 5.5 is now available:

  3. Grok 4.7 is now on Amazon Bedrock https://t.co/wM8bj1Wm9N

    Grok 4.7 is now on Amazon Bedrock https://t.co/wM8bj1Wm9N

  4. Today, with over 100 industry partners, we introduced the NVIDIA Open Agent Safety Platform, bringing together…

    Today, with over 100 industry partners, we introduced the NVIDIA Open Agent Safety Platform, bringing together OpenShell and Sentry. Artificial intelligence…

  5. Trained locally: ultra-fast 0.8B/2B System 1 decision models that match Jev on benchmarks and Doom, ~30 ms per decision (open weights)

    TL;DR: The Jeff models are a set of Qwen3.5 and Gemma fine-tunes for zero-shot classification: small, efficient, open-weight models with respectable…

  6. BA Computer Science, but fell in love with machine learning and AI. Just got my personal research accepted at NeurIPS as a poster. [R]

    I want to attend and present my findings in Atlanta. How is the vibe there? Are people overly critical or are people generally open-minded?

  7. Claude Sonnet 5.5 Released

  8. Finally found a model my hardware can run at full precision: me

    Was getting bored trying to squeeze every last t/s out of my local model on my hardware, so I made a tps counter for my fingers instead (with real tokenizers,…

  9. Reuters: Anthropic files for $2T IPO with $42B net loss in 2025, expects to spend half a trillion in 2027

  10. Nvidia wants to put a watchdog chip next to every AI agent including Claude, and Anthropic and SpaceXAI are on board

    This seems like quite a good solution. What do you guys think?

官方发布

  1. Z.ai GLM 5.3 (zai-glm-5-3) is now Generally Available.

    Z.ai GLM 5.3 (zai-glm-5-3) is now Generally Available.

  2. Introducing GPT-6 Sol and Luna

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

  3. codex rust-v0.159.0

    Release 0.159.0 (also released that day: 0.160.0-alpha.6, rust-v0.160.0-alpha.5, rust-v0.160.0-alpha.4, 0.160.0-alpha.3, 0.160.0-alpha.2, 0.159.0-alpha.13)

  4. Coding sessions are longer and use more context. Claude Opus 5.5 is built with that in mind.

  5. safety\_identifier request field

    You can now send safety_identifier, an opaque end-user identifier assigned by your application, on Chat Completions, the Responses API, deferred chat…

  6. Better prompt caching for GPT-6

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

  7. Giving companies more control over their AI agents, with NVIDIA

  8. Team Bots: AI coworkers that learn from your team

  9. Gemini 3.8 text-to-speech says hello

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

  10. Claude discovers a novel enzyme system with CRISPR-like repeats

    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…