AI Resonance · 阅读最新日报 · AI 入门推荐

2026-10-25 AI 日报

统计日时区：America/Los_Angeles · 尚未结算

makecindy/cindy · Zagitta/gpu-unlock · AgriciDaniel/claude-seo

在应用中阅读本期

开源项目

  1. makecindy/cindy

    Consider it done. The open-source AI agent that works out of the box · 想到，就能做到。开源、开箱即用的 AI Agent。

  2. Zagitta/gpu-unlock

    Unlock FP/BF16 instruction issue rate on Nvidia 3000's cards to their professional equivalents

  3. AgriciDaniel/claude-seo

    Universal SEO skill for Claude Code. 26 sub-skills + 19 sub-agents covering technical SEO, E-E-A-T, schema, GEO/AEO, agent readiness (Lighthouse Agentic…

  4. nightly-labs/openbot

    A local-first desktop workspace for persistent AI teammates. Run Codex, Claude, and Grok with dedicated workspaces, task queues, file sharing, browser…

  5. TauricResearch/TradingAgents

    TradingAgents: Multi-Agents LLM Financial Trading Framework

  6. Nutlope/hallmark

    Anti-AI-slop design skill for Claude Code, Cursor, and Codex.

  7. deepseek-ai/DeepInfer

    DeepInfer: High-performance, large-scale inference framework for DeepSeek models

  8. HKUDS/DeepTutor

    DeepTutor: Lifelong Personalized Tutoring. https://deeptutor.info/.

  9. microsoft/skill-recorder

    Desktop app that records your on-screen work session and uses the GitHub Copilot SDK to reconstruct it as an intent + ordered steps, then builds a reusable…

  10. Lordog/dive-into-llms

    《动手学大模型Dive into LLMs》系列编程实践教程

论文

  1. Beyond Gain: Measuring Skill Internalization for Agent Skill Retirement

    LLM agents retrieve skills from public hubs that hold up to about 100,000 entries and serve many models. As models improve, a hub must find the skills a given…

  2. FlexGlue: Sparse-to-Dense Feature Matching and Visual Disambiguation

    To estimate correspondences across images, existing approaches match either sparse interest points, which is highly efficient but limited by repeatability, or…

  3. A Tutorial on Physics Constrained Machine Learning for Hybrid Modeling, Optimization, and Control

    Physics-constrained machine learning (PCML) integrates domain knowledge, first-principles, and physical constraints into AI/ML models. In this work, we…

  4. DriveOSS: Dataset and Benchmark for 3D Anomaly Segmentation and Open-set Semantic Segmentation in Autonomous Driving

    Recent advances in 3D semantic segmentation have significantly improved scene understanding in autonomous driving. However, existing methods are typically…

  5. Flownomaly: Single-Step Conditional-Mean Recovery for Few-Shot Anomaly Detection

    Few-shot industrial anomaly detection requires modeling category-specific normality from only a handful of defect-free images. Common approaches tie inference…

  6. SceneBloom: Agentic Indoor Environment Generation with Realistic and Diverse Layouts for Embodied Learning

    Training generalizable embodied agents requires realistic, varied, and simulation-ready indoor environments. Language-guided methods support semantic planning…

  7. ReLEAP: Upcycling Vision Datasets through Cross-Task Unification and Semi-Supervised Annotation

    Developing robust vision models requires large amounts of annotated data, especially for tasks involving fine-grained class distinctions under variable…

  8. RootDown: Deploying and Benchmarking Models for Beet-Root Detection on Edge Devices

    Quality inspection of sugar beets after harvest and storage is critical to preserve sugar content and ensure optimal processing efficiency. The transfer of…

  9. PoKER: A Probability of Kill Estimation Rate Model for Air-to-Air Missiles Using Machine Learning on Stochastic Targets

    This work introduces PoKER, a novel probabilistic model engineered to optimize missile launch effectiveness in air-to-air scenarios, specifically within…

  10. Who Verifies the Tools? How Warped Formatters Mislead Agents and Their Verifiers

    Agents and verifiers often rely on the text a tool prints, and that text can misrepresent what the tool actually did, even when the tool's computation is…

行业新闻

  1. OpenAI and Anthropic cut their LLM prices by up to 80% within a few months. Why?

  2. A 10-year-old CNN outperforms frontier LLMs for skin cancer diagnosis

  3. IBM CEO says open weight models are now good and cheap on-prem

  4. PaperPilot – Zero-Cloud, Pure-Rust PDF Engine with WASM and MCP

  5. Attention Is Not What You Need (2025)

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

  6. R/open source bans all solo projects and projects using LLMs

  7. Microsoft Decision-1: 9B Qwen Model Launch Explained

  8. New Open-Source Cybersecurity Platform Uses AI Agents to Hunt Threats

  9. Show HN: Open-Source Cybersecurity Platform Uses AI Agents to Protect Businesses

  10. Ask HN: Are tab completion models dead?

    I think there is still place for this type of coding where you autocomplete predictable chunks and review it as you go. But I haven't seen much development in…

社区动态

  1. OpenAI and Anthropic cut their LLM prices by up to 80% within a few months. Why?

  2. Nanbeige releases Nanbeige4.3-33B, an Apache 2.0 teacher model for smaller agents [N]

    Nanbeige released Nanbeige4.3-33B as the teacher for its planned smaller agents, including an upcoming 3B model. the weights are Apache 2.0, and the model…

  3. qwen: merge tensors on convert by am17an · Pull Request #31028 · ggml-org/llama.cpp

    ...time to reconvert GGUFs?

  4. Nanbeige/Nanbeige4.3-33B · Hugging Face

    Stronger than Qwen3.8-27B across a range of benchmarks. but also lays the necessary foundation to power our upcoming small model distillation pipeline (e.g.,…

  5. RTX 3080 20GB got reBAR and P2P unlock! +70% prompt processing speed

    P2P is finally unlocked for RTX 3080 20GB - this leads to massive performance improvements in tensor parallel mode prompt processing speed. First, i want to…

  6. Backup your quants!

    source: https://huggingface.co/mradermacher/model_requests I'm sure the situation is the same for other quant producers. If you have anything you want to…

  7. CrowdGPT - A collaboratively trained LLM.

    Hello, I'm developing CrowdGPT, a project aiming to create an LLM without datacenters and without the usual millions of dollars spent by AI companies. The…

  8. I built a small validation layer that stops LLM agents from sending bad tool calls (TS + Python, MIT

    Agents call tools with arguments the model made up: an order ID that doesn’t exist, a negative amount, a date like “yesterday”. Most of us patch this with…

  9. What's the hardest part of creating edge cases for fine-tuning open-weight models? [D]

    I'm interested in how people build datasets for fine-tuning open-weight models, especially when it is about making a model handling edge cases than just…

  10. 🧵 Round 1 - DeepSeek API vs OpenCode Go vs CommandCode GOAT Finally here!

    🧵 Round 1 - DeepSeek API vs OpenCode Go vs CommandCode GOAT Finally here! This time, you're able to view the demos yourselves and judge.…

官方发布

  1. Z.ai GLM 5.3 Flash (zai-glm-5-3-flash) is now available.

    Z.ai GLM 5.3 Flash (zai-glm-5-3-flash) is now available. A multimodal 320B-parameter (18B active) Mixture-of-Experts model from Z.ai with a 1M-token context…

  2. The Decisions API is now generally available with gpt-6-luna.

    The Decisions API is now generally available with gpt-6-luna. Cache shared input across requests for prompt-cache discounts, include images in individual…

  3. Kimi K3.1 Pro

  4. Using thousands of AI agents to uncover new leads in biological data

    See how OpenAI, PerturbAI, and Scripps Research used distributed AI agents to uncover promising research leads in data from millions of brain cells.

  5. Advancing coding in low‑resource programming languages

    Learn how OpenAI is improving ChatGPT across low-resource programming languages, with new evaluations for software engineering, SQLite, and COBOL migration.

  6. Voxtral 3

    Voxtral 3: Mistral's voice agent SDK for real-time speech in 50+ languages — accurate transcription, emotional TTS, deep tool use. EU-hosted or self-deployed.

  7. Streaming comes to API Gateway: real-time and AI-ready APIs, without changing your spec

    API Gateway now supports real-time streaming—including Server-Sent Events, WebSockets, gRPC, and chunked transfer encoding—in Public Preview without requiring…

  8. Anthropic economic index connector

  9. Apple xcode claude agent sdk

  10. Native Model Authoring for Edge with LiteRT Tensor API

    Meet the LiteRT Tensor API: giving developers native C++ control over the model graph to unify end-to-end ML computations on heterogeneous edge hardware.