全生态物理 AI 落地，小米的主场攻防战

过去三年，大模型的快速进步，首先发生在数字世界。文本、图片、代码、视频都可以被转化为数据，模型通过海量预训练学习其中的规律，再完成理解、推理和生成。但数字世界能够覆盖的任务和场景终究有限。…

• AI 如果要进一步扩大产业价值，就需要从信息处理走向真实世界，开始理解环境，并与物理世界发生交互。
• 如果把这一轮 AI 浪潮分成两个阶段，上半场争的是模型，下半场的竞争则正在越来越多地转向物理 AI 的落地。
• 在今年规模最大的消费电子与未来科技展会 IFA 上，雷峰网也清晰地感受到：人工智能正在从生成内容、解决数字任务，进一步走进物理世界。

RSS 官方收录 · 可信分层展示

GPT-6 曝光， OpenAI 总裁说：AGI 来了

作者｜桦林舞王 编辑｜靖宇 他措辞很谨慎，用的是第一人称，加了「可能」，还专门留了后路说「如果你想说这是第一个，我认为这是合理的」。这不是 OpenAI 官方宣布 AGI 到来，而是一位公司总裁在镜头前说出了自己的判断。…

• 当地时间 9 月 3 日，GPT-6 Astra 正式发布。
• 极客公园拆解了目前能获取到的所有技术细节和演示材料，试图回答一个最核心的问题，这个模型，到底能干什么，能干到什么程度？
• 它真的能代表 AGI 的到来吗？

RSS 官方收录 · 可信分层展示

完成4000万美金融资，ACE成为AI音乐赛道融资额最高的华人团队

AI音乐公司ACE近日完成近4000万美元新一轮融资，由头部战投、CCV、顺为、Alphaist等机构投资。至此，ACE成为AI音乐赛道融资额最高的华人创业团队之一。本轮融资将主要用于音乐基础模型研发、全球团队扩张，以及面向大众市场的新产品Miya。…

• ACE正在打造一个AI-Native音乐平台：通过AI同时重塑音乐的生产与消费，成为AI时代的Spotify。
• 自研模型进入全球第一梯队ACE于2025年第一季度开始自研音乐生成基础模型。
• 经过一年多迭代，目前模型在主流评测指标上已领先绝大多数音乐生成模型，仅次于当前SOTA模型Suno v5.

RSS 官方收录 · 可信分层展示

Spring Boot 中的后量子密码学：一个冲刺周期内即可交付的四种模式

点击查看原文>

RSS 官方收录 · 可信分层展示

IDC首次发布零售云份额报告：腾讯云解决方案市场居首

9月4日，国际数据机构（IDC）发布《中国零售云市场份额，2025》 （以下简称为《报告》）。腾讯云通过整合PaaS、SaaS及CaaS（咨询即服务）能力，形成覆盖平台、应用及AI服务的产品矩阵，以23.…

• 4%的市场份额在解决方案市场位居第一。
• 同时，腾讯云2025年零售云收入同比增长16.
• 8%，在头部厂商中增速最快。

RSS 官方收录 · 可信分层展示

Cloudflare Wallets 迟来入局 x402：支出控制仅能约束单笔支付

点击查看原文>

RSS 官方收录 · 可信分层展示

Bioinfoysis Technical Report

arXiv:2609.…

• 03871v1 Announce Type: new Abstract: Large language model agents have …
• This design is poorly suited to long-horizon bioinformatics tasks, whe…
• We introduce \textbf{Bioinfoysis}, a multi-agent harness that represen…

RSS 官方收录 · 可信分层展示

STAIR (STructure Aware Information Retriever): A novel dataset and LLM based retriever for document structure augmentation

arXiv:2609.…

• 03874v1 Announce Type: new Abstract: Retrieval Augmented Generation (R…
• LLMs are improving at handling long context, but still suffer from "lo…
• Thus, precise and accurate retrieval is important.

RSS 官方收录 · 可信分层展示

Adapting to Evolving Requirements: Agentic AI for Retail Supply Chain Operations

arXiv:2609.03860v1 Announce Type: new Abstract: Retail supply chain operations rely on coupled decision modules that must adapt as requirements evolve.…

• LLMs offer a natural-language interface for this task, but existing me…
• Extending them to heterogeneous decision pipelines is challenging beca…
• We formulate requirement-driven adaptation as the joint selection of a…

RSS 官方收录 · 可信分层展示

Semantic Bayesian World Models

arXiv:2609.…

• 03834v1 Announce Type: new Abstract: Knowledge graphs describe reality…
• We argue that this mismatch is why the integration of language models …
• We envision Semantic Bayesian World Models (SBWMs): a Web that describ…

RSS 官方收录 · 可信分层展示

CauseCollab: Causal Unified and Modality-Agnostic Network for Heterogeneous Collaborative Perception

arXiv:2609.…

• 03818v1 Announce Type: new Abstract: Collaborative perception enhances…
• Recent protocol-based two-stage methods alleviate this problem by mapp…
• To address this issue, we propose CauseCollab, a causal unified and mo…

RSS 官方收录 · 可信分层展示

SVG-Score: Human-Aligned Evaluation of Text-to-SVG Generation

arXiv:2609.…

• 03806v1 Announce Type: new Abstract: Scalable Vector Graphics (SVG) ge…
• Progress, however, is held back by the lack of domain-specific evaluat…
• We introduce \textbf{\ours}, a human-aligned evaluation framework for …

RSS 官方收录 · 可信分层展示

Transfiver: Human-AI Co-Inference through a Shared Editable State

arXiv:2609.…

• 03797v1 Announce Type: new Abstract: Long-term human-AI interaction is…
• We introduce the TRANSparent Framework for Interactive, Verifiable, Ed…
• Its central idea is that interaction-specific information is maintaine…

RSS 官方收录 · 可信分层展示

Govern the Model, Not Only the Data: Storage, Circulation, and Learning in Creative AI

arXiv:2609.…

• 03800v1 Announce Type: new Abstract: Federated learning is increasingl…
• It borrows the vocabulary of the federated social web, yet inverts its…
• We argue that federation is not in itself a remedy for extractive AI, …

RSS 官方收录 · 可信分层展示

DNative-Twin: Decision Graphs and Digital Twins for Reconstructable Agentic Decisions

arXiv:2609.…

• 03787v1 Announce Type: new Abstract: AI agents increasingly gather evi…
• A final output alone cannot show which evidence, tool state, rule, aut…
• We present DNative-Twin, a graph-native digital twin that records a co…

RSS 官方收录 · 可信分层展示

Rethinking World Models for Safety-Critical Embodied Systems

arXiv:2609.…

• 03774v1 Announce Type: new Abstract: World models have progressed from…
• However, high predictive likelihood and visual fidelity do not necessa…
• This perspective identifies three structural mismatches in current wor…

RSS 官方收录 · 可信分层展示

千问办公上线首月用户数突破3000万，企业用户占比过半

点击查看原文>

RSS 官方收录 · 可信分层展示

SimSkill: A Lifelong Learning AI Agent for Autonomous Mastery of Traffic Simulation

arXiv:2609.…

• 03753v1 Announce Type: new Abstract: As large language models (LLMs) b…
• We introduce SimSkill, a self-evolving agent built around the Simulati…
• SimSkill identifies capability gaps, generates and solves environment-…

RSS 官方收录 · 可信分层展示

Proactive Service Agents: A Unified Decision Framework, Methods, and Evaluation

arXiv:2609.…

• 03727v1 Announce Type: new Abstract: Large language model agents can p…
• Proactive service moves the decision upstream: an agent must infer ser…
• This survey gives an operational definition centered on initiative and…

RSS 官方收录 · 可信分层展示

Artificial Intelligence for Energy Optimization in Data Centers

arXiv:2609.03716v1 Announce Type: new Abstract: Data centers are increasingly optimized by artificial intelligence and, at the same time, increasingly loaded by it.…

• The literature treats these as two unrelated problems: control studies…
• We screen roughly 194 papers retrieved through a documented protocol, …
• Of 28 primary control-oriented studies, 18 are validated in simulation…

RSS 官方收录 · 可信分层展示

Counterfactual Routing Using Integer Programming with Constraint Generation

arXiv:2609.03707v1 Announce Type: new Abstract: We present our submission to the IJCAI 2025 'Counterfactual Routing Competition' (CRC 25).…

• The goal of the competition is to find counterfactual explanations for…
• This requires deciding what the minimal changes to a road network woul…
• This enables explanations such as "Your suggested route would indeed h…

RSS 官方收录 · 可信分层展示

Synthetic Semantic Supervision for Contrastive Code Representation Learning in Small Transformers: An Empirical Study

arXiv:2609.03702v1 Announce Type: new Abstract: General-purpose code embeddings power tools for code search, classification, and retrieval.…

• Compact transformer encoders for code typically rely on either human-w…
• We empirically study an alternative: contrastive pretraining of small …
• We benchmark this approach against pretraining-based baselines, genera…

RSS 官方收录 · 可信分层展示

具身智能落地的最后20%，藏在「云」里

今年的WAIC和WRC，具身智能展台的风向变了：不再执着于炫技，转向务实。去年的画风还是十八般武艺，同台竞技——机器人跳街舞、格斗秀、走猫步等；而今年则是走进真实场景：机器人开始被放进物流、工业、家庭等具体任务里。…

• 过去一个多月，雷峰网一线走访了灵初智能、艾欧智能等具身智能公司，也看了各厂商在展台上的演示，一个感受越来越强烈：真实场景里的任务，远比Dem…
• 在走访和交流中，雷峰网也与多位云专家进行了交流。
• 当被问及机器人落地、商业化时，一位云专家表示，具身智能存在明显的“长尾效应”：前60%、70%、80%的进展可能很快，但最后20%却需要花非…

RSS 官方收录 · 可信分层展示

Analysis of Prompt Engineering for Drug Toxicity Prediction

arXiv:2609.03635v1 Announce Type: new Abstract: Clinical trials in the UK can cost up to {\pounds}1.…

• 3 million, with approximately 90% drug failure rate.
• Toxicity is a major contributing factor in drug failure.
• Testing is time and cost intensive.

RSS 官方收录 · 可信分层展示