AAI Search
AI 资讯日报 · 2026-08-25

2026-08-25 · AI 资讯日报

今日新收录 122 条公开资讯，按模型 / 产品 / 行业 / 论文 / 观点 自动归类汇编（非 AI 生成，点击可溯源原文）。

今日精选6 条

  1. 1.
    论文研究How to Train a Critic Stably and Efficiently— arXiv

    Group-based reinforcement learning methods such as GRPO for large language models avoid training a critic by sampling multiple responses for each prompt. A reliable critic could instead estimate token…

  2. 2.
    论文研究SWE Refactor Bench: Can Coding Agents Complete a Long-Horizon, Whole-Repository Stack Migration?— arXiv

    Modern software systems accumulate technical debt over decades of development, which makes migration expensive and largely manual. As coding agents become increasingly capable at bug fixing, can they…

  3. 3.
    论文研究EG-ARSA: An Expert-Grounded Open Model for Visual Road Safety Auditing in Low-Resource Settings— arXiv

    Road traffic injuries remain a major challenge in low- and middle-income countries, where proactive road safety auditing is limited by incomplete crash records, shortages of qualified auditors, and th…

  4. 4.
    论文研究Physics-Constrained Deep Learning Model for Contactless Blood Pressure Monitoring from Triaxial Bodyseismography— arXiv

    Ballistocardiography (BCG) is promising for unobtrusive long-term blood pressure (BP) monitoring in laboratory settings, but traditional BCG signals are vulnerable to the variations in body-bed intera…

  5. 5.
    论文研究Provably adaptive sampling with uniform and remasking discrete diffusion models— arXiv

    Discrete diffusion models offer a promising alternative to autoregressive generation by enabling parallel updates, but their sampling efficiency can depend strongly on the choice of the forward proces…

  6. 6.
    论文研究ConvergeFlow: Language Flow with Provable Convergence to Token Embeddings— arXiv

    Recent advances in continuous diffusion and flow-based language models (LMs) have achieved performance competitive with discrete LMs. However, existing continuous frameworks still rely on decoders sup…

模型发布/更新6 条

  1. 1.
    The full stack behind abundant intelligence— OpenAI

    OpenAI CFO Sarah Friar explains how advances across chips, compute, models, and products compound to deliver more useful intelligence at greater scale and lower cost.

  2. 2.
    Jalapeño’s first results show industry-leading speed and efficiency in AI inference— OpenAI

    Jalapeño is a custom inference chip from OpenAI that delivers faster, more power-efficient AI inference, with higher throughput and lower latency for modern models.

  3. 3.
    Google launches Gemini for legal work to automate contracts and research— The Decoder

    AI 导读 · 谷歌切入法律垂直场景，用MCP连接器打通主流工具，企业级AI应用再下一城。

  4. 4.
    Funding better evaluations of AI’s impact on wellbeing— Anthropic

    AI 导读 · AI福祉影响评估获资金支持，伦理监管短板有望补强。

  5. 5.
    Disrupting a new covert influence campaign from Russia— OpenAI

    OpenAI banned Russia-origin accounts using AI to promote a fake Israel-based think tank and a “sovereignty” index praising Russia and criticizing the West.

  6. 6.
    Introducing the Admin plugin for ChatGPT Work and Codex— OpenAI

    Use the Admin plugin for ChatGPT Work and Codex to analyze workspace usage, manage members and permissions, adjust limits, and act on admin requests.

产品发布/更新6 条

  1. 7.
    Zyrexnn/Cybermes— GitHub

    Autonomous Offensive Security, Bug Bounty & Red Teaming Agent Framework powered by Hermes Agent, specialized reasoning skills, and multi-model LLM orchestration…

  2. 8.
    Gnosil/semantix— GitHub

    semantic agent kernel which make agent efficient and self-evolve

  3. 9.
    开源国产8B模型，比肩闭源Image 2了！— 量子位

    SenseNova U1.5 Lite

  4. 10.
    AI4S开始进入「项目时代」：紫东太初把AI从做Task推向做Project— 量子位

    AI科研能力从工具级转向系统级

  5. 11.
    WAIC CONNECT | AI出海，不聊虚的！CONNECT带你拿下马来西亚真正的AI采购需求— 量子位

    2026年9月7日–8日，WAIC CONNECT MALAYSIA活动，将在吉隆坡盛大开启！

  6. 12.
    具身大满贯还全开源！原力灵机DM0.5登顶RoboDojo，且clone且珍惜— 量子位

    机器人大脑新SOTA

行业动态6 条

  1. 13.
    How much of HN is AI?— Hacker News

    AI 导读 · AI含量决定社区价值，这份数据透视值得深挖。

  2. 14.
    Bookshelf – Self-hosted eBook library that runs on object storage— Hacker News
  3. 15.
    Gamma acquires Accel-backed design startup Lica— TechCrunch

    AI 导读 · 设计人才流向AI产品，预示Gamma正加速布局生成式设计的底层创新。

  4. 16.
    OpenAI’s Jalapeño chip is built for fast inference at scale, benchmarks show— TechCrunch

    AI 导读 · 推理性能与能效双突破，为规模化AI部署提供新标杆。

  5. 17.
    Meta's paid AI agent Hatch launches soon, with a new model called Watermelon due in October— The Decoder

    AI 导读 · Meta加码商业化AI，Hatch与Watermelon组合拳值得关注。

  6. 18.
    Accel-backed Keenable is indexing the web for AI agents— TechCrunch

    AI 导读 · AI代理专用搜索索引获2600万美元融资，瞄准下一代交互入口。

论文研究6 条

  1. 19.
    How to Train a Critic Stably and Efficiently— arXiv

    Group-based reinforcement learning methods such as GRPO for large language models avoid training a critic by sampling multiple responses for each prompt. A reliable critic could instead estimate token…

  2. 20.
    SWE Refactor Bench: Can Coding Agents Complete a Long-Horizon, Whole-Repository Stack Migration?— arXiv

    Modern software systems accumulate technical debt over decades of development, which makes migration expensive and largely manual. As coding agents become increasingly capable at bug fixing, can they…

  3. 21.
    EG-ARSA: An Expert-Grounded Open Model for Visual Road Safety Auditing in Low-Resource Settings— arXiv

    Road traffic injuries remain a major challenge in low- and middle-income countries, where proactive road safety auditing is limited by incomplete crash records, shortages of qualified auditors, and th…

  4. 22.
    Physics-Constrained Deep Learning Model for Contactless Blood Pressure Monitoring from Triaxial Bodyseismography— arXiv

    Ballistocardiography (BCG) is promising for unobtrusive long-term blood pressure (BP) monitoring in laboratory settings, but traditional BCG signals are vulnerable to the variations in body-bed intera…

  5. 23.
    Provably adaptive sampling with uniform and remasking discrete diffusion models— arXiv

    Discrete diffusion models offer a promising alternative to autoregressive generation by enabling parallel updates, but their sampling efficiency can depend strongly on the choice of the forward proces…

  6. 24.
    ConvergeFlow: Language Flow with Provable Convergence to Token Embeddings— arXiv

    Recent advances in continuous diffusion and flow-based language models (LMs) have achieved performance competitive with discrete LMs. However, existing continuous frameworks still rely on decoders sup…

技巧与观点6 条

  1. 25.
    5 ways to upgrade your home decor with Google Search— Google AI

    AI 导读 · 搜索引擎跨界家居场景，实用工具挖掘生活灵感，别错过这波AI落地新姿势。

  2. 26.
    Granite 4.2 LLMs: How They're Built— HuggingFace Blog

    AI 导读 · 看点：IBM开源模型技术细节全公开，看企业级AI如何平衡性能与成本。

  3. 27.
    Quantization-Aware Healing: a compressed, 4-bit model that outperforms its full-precision original— HuggingFace Blog

    AI 导读 · 量化修复让4位模型反超原版，颠覆压缩必然降智的行业共识。

  4. 28.
    Wire It, Run It, Deploy It: AI Workflows in Gradio— HuggingFace Blog
  5. 29.
    not much happened today— AI News

    **Agent harnesses** are becoming a key optimization focus, with NVIDIA research showing traditional skill checks poorly predict agent usefulness and proposing a new metric called *…

  6. 30.
    How Hugging Face Inference Endpoints, Jobs, and Buckets Power Search on Papers with Code— HuggingFace Blog

快讯

日报生成时间：2026-09-06
← 2026-08-24最新日报2026-08-27 →