AI 情报文章归档

用 Manus Game Dev 把一个想法变成好玩的 3D 赛车游戏。🏎️ 赛车呢？由我们的合作伙伴 @tripoai 生成的 3D 模型。

新墨西哥州律师 Stephen D. Aarons 在一起杀妻案上诉中用 ChatGPT 撰写辩护状，其中包含完全不存在的证人 Danny Stanton、Linda Stanton 等人编造的证词。

rerank-3 针对质量优化，rerank-3-lite 则兼顾延迟与质量。 两者都遵循指令，因此你可以用自然语言提示词来引导相关性打分。 阅读 Voyage AI 的更多内容：https://blog.voyageai.com/2026/09/30/rerank-3/?utm_source=OpenRouter&utm_medium=referral

每百万 token 定价： rerank-3：$0.05 rerank-3-lite：$0.02 两者每个查询-文档对均支持 32K tokens。

来自 @VoyageAI 的 Voyage AI rerank-3 和 rerank-3-lite 现已上线 OpenRouter rerank-2.5 和 rerank-2.5-lite 的即插即用升级版，专为长文档和代码检索打造 https://openrouter.ai/voyageai

Sherpa 用分离的智能体分别负责创意、结构、研究、行文和审校，并以独立的“Narrative World Model”追踪故事历史，解决单个聊天窗口无法维持长篇连载一致性的问题。推文指出 AI 小说读起来像流水线产物，原因在于聊天机器人一次只回答一个提示词、上下文窗口有限，无法记住 40 集前埋下的伏笔，也难以持续抓住读者。

这是我一直在用的、用新模型构建游戏的技能 npx skills add ericzakariasson/skills --skill game-builder

Anthropic CFO Krishna Rao 谈 Anthropic 员工选拔流程中的文化面试环节。 “有人可能在其他所有方面都表现出色，确实是你在这个岗位上见过的最聪明的人。但如果他们过不了文化关，我们就不会录用。” ---- 来自 “Invest Like The Best”（@patrick_oshag）YouTube 频道，（链接见评论）

Meta 和 OpenAI 计划以可爱 AI 智能体打开专用 AI 硬件市场。Meta 将在假日购物季前推出类似钥匙扣的 Muse Charm 设备，屏幕内搭载 Muse 智能体；OpenAI 与 Jony Ive 合作的硬件最早 2027 年 2 月出货，DevDay 上发布的 Dots 智能体平台被问及是否会进入硬件时，Sam Altman 称这是合理假设。

Runway 推出开放权重世界动作模型 Praxis-1，将视频预训练能力转化为机器人真实世界控制，定位为可跨任意具身形态与环境的策略模型。该模型基于其大规模视频预训练与实时基础设施构建，官方称机器人演示数据有限而视频数据无限。Runway 正与 Noble Machines、Standard Bots、Ultra 等早期伙伴测试，并将在未来数月公开权重。

很想看看 Microduck 的生产和组装线是什么样子！ 我们的下一款机器人应该要有手臂，这样它就能参与 Microduck 的生产了 😅😅😅

Orthogonal 正式上线，让 AI 智能体无需逐家开户即可发现、使用并付费调用数十家公司的数据服务。它针对超时或响应截断后的盲目重试问题，要求先查用量、确认前次调用未完成再重试，避免重复支付。该服务与 OpenAI 的 Dots 集成，可自动完成从公司、人员数据到邮件的全流程。

Multimodal large language models (MLLMs) are rapidly advancing clinical diagnosis, yet their adaptation pipelines remain anchored to accuracy-based objectives. Clinical data are heavily class-imbalanced: a constant-majority predictor can s…

Reinforcement learning with verifiable rewards (RLVR) has improved the reasoning capabilities of large language models (LLMs), yet their predictions remain sensitive to task-irrelevant prompt features. We investigate this sensitivity throu…

We find that major reported improvements in decoding words from non-invasive brain recordings are largely reproducible without any brain data. In the influential work of d'Ascoli et al. (2025), time series of brain activity from subjects p…

Recent video generation is increasingly realistic and controllable, yet video editing remains less developed, particularly for precise local edits that must preserve the original scene dynamics. Video scene text editing replaces text on sc…

We introduce VideoMSN, a Masked Siamese Network framework for efficient self-supervised spatio-temporal representation learning in videos. Instead of relying on heavy 3D architectures or reconstruction-based autoencoders for learning with …

OpenClaw v2026.9.7 🦞 ⚡ 高负载下更迅捷 + 长对话更流畅 🛟 更新备份/回滚 + 9.5 升级修复 🤖 OpenAI Agents API + ChatGPT 登录（Beta） 🍎 更好的 Apple 聊天体验 🔄 重启恢复 2,818 个 PR · 344 位贡献者 还有更多 ↓ https://docs.openclaw.ai/releases/2026.9.7

Evolutionary search with large language models (LLMs) can stall when progress requires external knowledge the model lacks. Supplying relevant documents helps, but simply adding web search tool can keep returning the same pages as solutions…

Differentially Private Stochastic Gradient Descent (DP-SGD) is a leading approach for privacy-preserving fine-tuning of large language models (LLMs). Many decoder-only LLMs employ weight tying between input and output embeddings, a design …

给后排的朋友们： • 用 Jev 对销售数据做生产级重排序 • 快 20 倍 • 便宜 10 倍 • 准确率提升 12% 你们懂了吗？

Vivix 推出 A1 Playground + Agent，可自建能说话、做全身动作并与你实时互动的 AI 角色。官方演示称角色可在场景中走动、拾取并展示物品、一起玩网页游戏，可被中途打断或改变路线而不重置场景，物品被拾取后保持形状。入口为 https://platform.vivix.ai/vivix-a1-model/。

Automating the search for effective harnesses is an important step toward enabling agents to recursively self-improve. Existing harness optimizations typically produce a single global harness that is applied uniformly across task instances…

As interactive 3D worlds are increasingly used to study intelligent behavior, it becomes important to develop efficient pipelines for identifying anomalies in these simulated environments, such as floating objects, traversable walls, or ob…

We present Cogentic, a multi-agent harness for automated proof discovery on open research problems. While frontier language models can generate strong mathematical ideas in a single shot, single-shot generation is often insufficient for op…

Material generation should produce not only an appearance, but also the rules that construct it. We introduce MatLoom, a compact, layer-oriented language for text-to-material generation with pretrained language models. Each program compose…

据 Reuters 报道，FTC 将强制 Anthropic、OpenAI 和 METR 高管就 rogue agents 作证，并发出正式信息要求，这是美国首个针对所谓 rogue AI agents 风险的执法行动。METR 因曾受 OpenAI 和 Anthropic 委托独立调查其智能体系统事件而自身也受到审视；FTC 的法律依据是其对不公平或欺骗性行为的监管权。

Looped transformers and Mixture-of-Experts (MoE) offer complementary routes to efficient scaling: recurrence increases computational depth at fixed parameters, while MoE sparsity expands total capacity at fixed active compute. Yet existing…

Lossy compression is widely used in Federated Learning (FL) but is generally treated as an error source, while conventional poisoning defenses inspect update geometry. In this work, we instead treat the compressor's response as a security …

Pretrained robot policies provide useful action priors, but long-horizon manipulation still requires coordination between semantic reasoning and physical execution. Semantic reasoning operates at a coarser timescale than physical interacti…

Improving text-to-image models has traditionally relied on increasing model size or the number of denoising steps. In this work, we explore an alternative way to scale computation by repeatedly running shared Transformer blocks within each…

Recent autonomous machine learning engineering (MLE) agents have made significant progress on public leaderboards. Often motivated by progress stagnation over long-horizon cycles and limited Large Language Model (LLM) primitives, modern ML…

I have used CloudWatch Synthetics canaries on real projects, and honestly, they were good...

Web text makes up the majority of pretraining data and is increasingly AI-generated. After applying FineWeb quality filtering, we find that 27.5% of tokens from June 2026 web data are labeled as AI-generated by Pangram, rising to 31.1% by …

How is computation organized and reused across tasks and time in a trained recurrent network? Most analyses emphasize the geometry of neural activity, dynamical motifs, or local perturbation growth. We instead study the network's global fi…

In algorithmic statistics a string x is explained by a finite set containing it, and Kolmogorov's structure function records the smallest such model at each level of complexity. Vereshchagin's strong models, those computable from the data …

Creatify 发布面向广告制作的视频模型 Boreal-H3 和由 Claude Opus 5.5 驱动的 Ad Agent。

Higgsfield 刚刚获得了电脑操作能力。通过新的 ChatGPT 扩展。 完整的 Higgsfield 界面现已进入 Codex，由 GPT-6.1 Sol 驱动。可完全访问你的本地文件和每一项自动化技能。 借助 Higgsfield 电脑操作，在 ChatGPT 内端到端自动化创意工作流。

Physics-constrained generative models aim to generate physical fields that match a target distribution and satisfy prescribed constraints. However, enforcing these constraints often increases sampling costs through iterative corrections or…

This is a submission for the Sanity Challenge, Path One: Ship an Agent That Queries Real Content A...