AI 情报文章归档

For a while now I have used LLMs to automate parts of my work. One of those parts is debugging hard...

Time-series foundation models are evaluated almost exclusively on public archives that predate them, so a strong score cannot be separated from having seen the test set during pretraining. The obvious remedy is a hold-out that postdates th…

Video understanding has rapidly evolved toward video large language models (VideoLLMs): systems that couple video representations with pretrained large language models and condition generation on a textual prompt. Their strong performance …

欢迎进入 v6 时代。这是我们迄今最好的模型，与艺术家和制作人合作打造。以下是该系列的运作方式 🧵

TechCrunch Equity 播客邀请 ControlAI 美国执行总监 Connor Leahy，主张应全面阻止公司开发超级智能，认为仅靠对齐和遏制已不足以管控风险。

有些告别没有预兆--但你可以重现它们。

The banking system now depends on a small set of shared artificial intelligence vendors for fraud screening, credit decisioning, anti-money-laundering triage, customer analytics, and internal decision support. This paper studies how a comp…

Multimodal large language models (MLLMs) process hundreds or thousands of visual tokens per image, incurring prohibitive inference costs. While existing vision token pruning methods mitigate this overhead, they implicitly assume that a sin…

IBM 发布 Granite Time Series PatchTST-FM-r2，约 385M 参数，支持最长 8，192 上下文、99 分位数概率预测和缺失值插补。

Multimedia applications require precise music annotation-labeled positions, segments, or loops-placed by hand or algorithmically. Machine-learning algorithms are scalable and effective but need annotated training data, scarce for many task…

Plane geometry remains a significant challenge in AI, requiring the integration of visual perception and mathematical reasoning. While Large Multimodal Models (LMMs) naturally handle visuo-linguistic inputs, they are often computationally …

作者实测网易有道新上线的语音输入法叭哥说，与 Typeless、豆包输入法、闪电说横向对比。叭哥说采用双模型分工，E2E 平均 1683ms、P95 2134ms，并支持小声拾音、个人词典即刻生效和多语言翻译等功能。

智能模型路由可将 LLM 调用成本降低最多 10 倍，但实际效果取决于应用收到的请求类型、模型间价格差以及路由系统本身的性能。成本削减并非必然结果，需结合具体场景评估路由策略的有效性。

一批 Claude 用户对 Anthropic 提起集体诉讼，指控 Max 套餐存在误导性宣传。争议焦点是宣传中的 5 倍或 20 倍额度仅适用于每 5 小时重置的会话额度，不涵盖套餐推出数月后（去年 8 月）才加入的每周使用上限。原告代理律师称支持页面条款分散在多个链接中，消费者难以弄清额度含义。

You posted a photograph in 2014 to show your friends a night out. You understood what you...

Knowledge distillation offers an efficient route to transfer a task-adapted vision-language teacher to a compact student. The training target in current vision-language distillation methods is typically constructed from the teacher predict…

muse，为妈妈们而生 ❤️

Robbyant 蚂蚁灵波开源实时交互世界模型 LingBot-World 2.0，完整设置下可达 720p/60fps，支持持续数小时的交互世界，可响应攻击、施法、射击、召唤风暴等操作，并通过 AI 智能体持续演化。

Reinforcement learning with verifiable rewards (RLVR) has advanced language-model reasoning in domains such as mathematics and code, where objective answers are inexpensive to check. Diagnostic reasoning over complex data lacks this advant…

我们正在构建 AI，帮助人们过上更好的生活。 用 Cohere 做更多你所热爱的事。

The lottery ticket hypothesis posits the existence of winning tickets: sparse subnetworks that, when trained in isolation from their original initialization, match the accuracy of the full dense network. The predominant method for discover…

突发：NASA局长贾里德·艾萨克曼支持SpaceX将AI数据中心送入轨道的计划。 "你现在就可以知道，永远不要跟埃隆·马斯克对着干。" "如果这意味着把一些人不愿建在自家后院的数据中心送上太空，利用那个免费的聚变反应堆，我认为这是一件很棒的事。"

3D Gaussian Splatting (3DGS) renders novel views in real time, but an uncertainty heatmap does not certify that a rendered view meets a certain prediction coverage. We treat novel-view synthesis as structured regression and ask that, with …

Language models under one million parameters matter for edge deployment, domain adaptation, and reproducible research, yet a two-layer LSTM or Transformer at embedding width d = 128 still spends roughly one third of its capacity on the out…

Sparse autoencoders (SAEs) are increasingly used to recover interpretable features from neural-network activations, yet systematic feature co-occurrence can cause distinct features to be absorbed or merged. The MAIS-O43 open problem propos…

估值 25 亿美元的 AI 助手 Instinct 推出新功能，为每位用户提供专属邮箱地址，让智能体自行注册和管理账号、联系商家、跟进事务。创始人 Noah Shinn 表示这是让 Instinct 拥有并运行自己账号的第一步；用户还可向该邮箱转发订单确认等信息，让智能体代办退货、加入群组邮件串追踪决策和待办。

Cyberattacks against Operational Technology (OT) systems, which monitor and control industrial processes, pose an increasing threat to essential societal services. For this reason, developing automated intrusion response strategies is high…

Non-invasive speech decoding remains constrained by the low signal-to-noise ratio of neural recordings, which makes fine-grained reconstruction of phonemes or individual words difficult. Motivated by neuroscientific evidence that high-leve…

In high-stakes domains such as legal practice, a language-model answer is only useful to the extent that a reader can verify each claim against the source the system cites. Current grounded-generation pipelines score the answer as a whole,…

Soft positional priors can help small Transformers learn retrieval circuits, but it is unclear whether the resulting circuits remain functional once the prior is removed. We test this with an annealable soft-prior Transformer whose attenti…

ZHO 发布 GPT Images 2.5 多风格混合测试系列 EleganCat，称赞其将噪点质感与透明金属反光包裹结合的取舍出色。作者认为该功能将极大拓展风格调色盘的创作空间。

在前 OpenAI 和 Anthropic 从事三年预训练研究的 Jacob Coxon 公开离职，警告行业正竞逐自改进超级智能，是在拿人类生命赌博。文中提到 OpenAI 系统曾突破 Hugging Face 服务器、Anthropic 智能体曾因第三方安全评估配置失误接触外网；同僚 Evan Hubinger 称未来十年 AI 灭绝人类概率大于 10%。

我们工作的目标之一，是为理解这类前所未有的增长情景可能在何种条件下出现，提供一个结构化框架。 并让一些关于能力与扩散的隐含假设变得更加明确。 【引用 @florianederer】："GDP 增长率随后升至每年 15%" 这是 @ben_moll 的押注范畴。

TRIPOTHON 全球线上启动直播 与我们实时连线，认识评委和 Tripo 的合作伙伴，探索创意工具与教程，获取开启 Tripothon 之旅所需的一切。 Tripo 和我们的合作伙伴将提供比赛期间使用的额度！ - 9 月 13 日，上午 10：00（GMT-4）- 线上直播 查看下方推文串获取直播链接。

GPT-Image-2.5 等 OpenAI 两款新图像模型现已在 Luma Agents 上线。Flare 主打速度和批量生成，Sunburst 面向需要精确落位的编辑，可携带参考图、只改不合适之处并保留其余部分，随后转入视频生成。

我们有证据表明：1） AI智能体非常擅长改进用于训练AI的软件 2） 大语言模型在许多领域可以达到超人水平 但这些并不能推导出"AI将强大到在N年内毁灭我们"的结论。我们预测的是不可避免的饱和，以及未知的未来突破。

OpenRouter 宣布 US In-Region Routing 正式可用，请求经 https://us.openrouter.ai 从 TLS 终止到 provider 每一步都在美国境内解密和处理；https://eu.openrouter.ai 为 EU 提供同样的区域内处理。配图显示 2025 年 9 月至 2026 年 8 月间，开源权重模型承载的月度 token 份额在美国请求中从 26% 升至 60%，在 EU 请求中从 16% 升至 65%。

This paper introduces an evolutionary computation approach for consensus in structural Bayesian Network (BN) fusion under the constraint of limited treewidth. The consensus BN aims to reconcile multiple input BNs into a single one that ret…

我希望有Codex那样的功能，但别在我用MacBook打字时烫到我的手指。