AI 情报文章归档

我们去年12月在伦敦开设了办公室。上个月，我们让一屋子的人聚在一起，正是他们让这一切变得值得。

Meta 在 Facebook 推出 AI 功能 Your Algorithm，用户可用自然语言描述偏好，影响 Feed 和 Reels 的内容推荐，调整实时生效。

我们的 CEO @MikeyShulman 入选了 @TheAnkler 的首届 Future of Entertainment 25 🎶 “Mikey 预见到一场重大变革即将到来，在音乐以及所有其他形式的娱乐中，‘人们将从被动消费内容转向主动创作内容’。” 阅读完整名单 👇 https://theankler.com/the-future-of-entertainment-25-new-stars-studios-storytellers/#tech

非常激动地欢迎 Soffia 和 Caco 加入 Conductor！ 我们最初是在 Discord 上发现 Faces 团队的。他们是理想的早期用户——他们喜欢 Conductor，但也毫不忌讳地告诉我们哪里不好用 ;) @SPLehman 注意到了他们的消息，私信了他们，我们一起吃了午饭，后面的事就顺理成章了。现在我已经无法想象团队没有他们会是什么样。全员上船！

阅读更多：https://devin.ai/blog/devin-in-pylon

通过我们的新集成，Devin 和 Pylon 保持同步，让 CSM 和工程师能够从 bug 报告到合并修复 PR 全程协作使用同一个智能体。

今天就在 https://devin.ai/integrations/pylon 试用

客户不再满足于快速响应。他们想要的是问题被快速修复。人类在工具之间复制粘贴上下文会拖慢你的速度。

推出 Devin in Pylon：加速你的中位修复时间（MTTF），新 Devin 客户可获 15,000 美元额度。

Google 开始全球推出 Gemini 4 Argon，称其为最强前沿模型，可通过 AI Studio、Gemini 应用和 Google Antigravity 使用；免费用户和 Plus 订阅者只剩较弱的 Flash 模型。

Gemini 4 Argon 正式进入 GA，在 Artificial Analysis Intelligence Index 上与 GPT-6 Astra 排名持平，Cost per Task 为后者的 60%。

立即体验：https://openrouter.ai/liquid/lfm-2.5-24b-a2b

Liquid AI 发布 LFM2.5-24B-A2B，为 LFM2.5 系列最大成员，现已在 OpenRouter 上线。该模型为 24B 总参数、每 token 2B 激活的 MoE 推理模型，上下文扩展至 131K tokens，词表 128K，使用 33T 预训练 tokens，面向笔记本和 PC 本地运行 Agent 与工具调用多步任务；在 agentic 和 tool-use 任务上可与总参数和激活参数多约 50% 的 Qwen3.6-35B-A3B 竞争，开源…

Databricks 首次进入 Gartner 分析与商业智能平台魔力象限 2026，获评"远见者"，并取得所有受评厂商中"愿景完整性"最高分之一，成为该报告 20 多年历史上首发表现最好的厂商。

The Verge 评测 Ray-Ban Meta Audio 音频智能眼镜，售价 $349，仅保留开放式音频功能并移除摄像头，佩戴者不再有被旁观者抵触的压力。

Agentic robotics places a large language model (LLM) agent in the robot control loop, where it composes perception, motion, and vision-language-action (VLA) tools under the guidance of a memory and thereby generalizes to new tasks without …

Sparse-to-dense motion completion recovers dense mesh motion from sparse vertex displacements. Residual priors trained on one geometry family frequently fail to beat parameter-free harmonic extension off-family. Moreover, cross-family trai…

How should demonstration quality be used during vision-language-action (VLA) pretraining to improve performance after target-task fine-tuning? We compare weighting, filtering, and conditioning across reward definitions and label granularit…

Long-horizon agents have drawn growing attention for their ability to work autonomously -- from hours to days -- on economically valuable tasks, a capability essential to automated discovery and recursive self-improvement (RSI). Yet open-s…

Cat Jam（隐身模型）现已在 OpenCode 中免费开放。官方称该模型速度非常快，专为智能体编码构建，并提供零数据保留。

Sampling equilibrium field configurations remains a major computational bottleneck in lattice field theory, motivating the development of generative samplers that can produce independent configurations. Diffusion models are promising for t…

Reasoning models achieve strong performance through long chains of thought (CoT), but often spend more tokens than needed to solve a problem. Existing ways to shorten the CoT require additional training for each model, add a fixed steering…

How can we build an agent that thinks beyond human language, yet still acts through it in the external environment? Modern LLM agents still reason and interact through text, while latent reasoning offers an efficient internal medium but la…

Generating information-rich visual artifacts such as slides, posters, and dashboards as code is appealing because the outputs are structured, editable, and verifiable. Yet code models cannot see how their choices of layout and color come t…

Persistent homology summarizes the topological evolution of a graph as persistence diagrams (PDs). Generating graphs whose topology matches a target descriptor is increasingly important, yet existing topology-aware generative models rely o…

Reinforcement learning with verifiable rewards (RLVR) relies on costly model-generated responses, motivating larger rollout batches and square-root learning-rate scaling. Reported to approximately preserve progress per sample with one opti…

Agor 开放托管版 Open Beta，一年前由 @mistercrunch 为追踪自己在多个终端的智能体会话而创建。产品把现有智能体 harness（如 Claude Code、Codex、Gemini）放在空间化看板上，会话和 git 分支按区块分组；用户可按任务自选智能体和模型，通过共享 Knowledge 和显式交接传递上下文。

Multi-objective molecular optimization is a critical process in drug discovery that aims to modify a given molecule to simultaneously improve multiple properties while maintaining a similar overall structure. This process often requires it…

Earth observation (EO) models are still built by hand. For each new task, experts design a training recipe and tune it against a benchmark. We show that this loop can be handed to an agent. Any verifiable task can be cast as a harness and …

Robot policies are expected to generalize across deployment environments, including variations in the robot itself: cameras are bumped and remounted, calibrations drift, controllers run at different rates, and end-effector pads wear down. …

Robots are increasingly autonomous in task execution, yet their improvement remains largely human-driven. This study investigates Physical Recursive Self-Improvement (Physical RSI), in which physical experience guides lasting updates to bo…

GitHub 公开其新密钥扫描分类器的设计与训练细节：先用正则生成候选，再用微调 ModernBERT 结合文件路径、变量名和上下文判断是否为真实密钥。相比此前 LLM 管线，仓库所有者加权接受率提升 11.5%，post-push 扫描中位执行时间降低 86%，服务成本下降超过 98%；模型导出为 ONNX 并做 8-bit 量化，且已将通用密钥检测扩展到 push protection。

This paper investigates offline Q-learning for finite-horizon Markov decision processes with Kullback-Leibler (KL) regularization. Recent studies have shown that KL regularization can yield fast rates in offline decision-making problems. H…

A central challenge in scaling flow models is maintaining stability in training and integration. In flow matching, this stability depends critically on how source samples are paired with data. Independent random pairing introduces substant…

It is difficult to accurately model the physics of the interaction between a robot and the objects it manipulates, including contact mechanics, coupled contact dynamics, deformations, stiffness and friction. We present RoboSimStudio (RSS),…

With the rapid development of multimodal large language models (MLLMs), evaluating their video understanding capabilities has become increasingly important. Existing video understanding benchmarks cover only part of the evaluation dimensio…

Simulation-based Bayesian inference for Chemical Reaction Networks (CRNs) must simultaneously solve two computationally demanding problems: discriminating between overlapping reaction pathways and estimating reaction rate constants. Classi…

也来看看 @nuvalab 上的 FastH3-Omni 支持 他们已经在交付面向生产级用例的 FastH3-Omni 预览版！ 敬请期待 10 月 20 日的完整开源发布！🚀🚀 https://x.com/nuvalab/status/2110062139872600106

Large language models (LLMs) generally rely on sampling temperature to regulate the amount of randomness in their outputs. Mainstream methods for selecting temperature typically involve fixed, heuristic global settings and context or token…