2026年11月14日 技术热点总结

项目简介：Apex Neural Rendering (ANR) by Apex Compute

技术栈：C++

核心介绍：Apex Neural Rendering by Apex Compute. 1. Download a package from Releases: 2. Close the game, extract the package and run the installer with the folder of the game’s .exe: prints: WINEDLLOVERRIDES=”dxgi=n,b” %command%, or winmm=n,b in Vulkan games 3. In the game, turn on frame generation (the NVIDIA option where the game offers several). To uninstall, run uninstall.cmd or sh uninstall.sh with the same folder. On Windows 10 and 11, and on Linux and SteamOS with Proton.

项目数据：⭐ 1,467 Stars，🍴 21 Forks

项目简介：Open-source, self-hosted Slack alternative where your team and AI agents work together in channels and DMs. Agents run on Claude, ChatGPT/Codex, Gemini or Grok, with MCP tools, their own computer and browser, routines, memory and approvals. Web, desktop and phone apps. Docker Compose + Postgres. AGPL-3.0.

技术栈：TypeScript、agentic-ai、ai-agents、browser-automation、chat、claude、claude-code、codex、collaboration

核心介绍：agentes.chat The open-source, self-hosted Slack alternative where your team and AI agents work together. Website · Cloud · Self-hosting · Features · Discussions A…

项目简介：Private AI meeting notes that stay on your computer. A free, open-source alternative to Granola for Mac, Windows and Linux.

技术栈：TypeScript、ai-meeting-assistant、ai-notetaker、electron、granola-alternative、linux、llama-cpp、local-ai、local-first

核心介绍：Private AI meeting notes. A free, open-source alternative to Granola. Download for Mac · Windows · Linux Yogurt records your calls without adding a bot to the meeting. macOS 14.4+ (Apple Silicon) · Windows 10 2004+ (x64) · Ubuntu 22.04+ or similar (x64 AppImage) · 8 GB memory yogurt.so · Help · Build and run locally · MIT license M…

项目简介：One self-hosted app for your shows, movies and music.

技术栈：Go、docker、homelab、jellyfin、media-server、plex、self-hosted

项目数据：⭐ 1,099 Stars，🍴 57 Forks

项目简介：Renders an r/homelab-style network diagram from a YAML inventory (Python → SVG/PNG). Written with Claude.

技术栈：Python

核心介绍：The script behind the network diagram of my homelab that I posted on r/homelab. It turns a YAML file describing the lab into a big SVG/PNG diagram in the style of the diagrams you see on r/homelab.

项目数据：⭐ 985 Stars，🍴 82 Forks

项目简介：Skills for human-led AI teams: delegate tasks, verify evidence, explain steps clearly, and learn from real cases. 一个人统筹 AI 团队：派活、核证、清楚说明、经验复用。

技术栈：agent-skills、ai-agents、collaboration、human-in-the-loop、workflow

核心介绍：仓名：solo-to-team-skills 英文副标题：Skills for human-led AI teams. 中文说明：一个人统筹 AI 团队的协作技能包。 署：大冰发起 一个人，也能带起一支 AI 团队。 这是一套拿得到、装得上、叫得动的协作技能包：四支技能＋一个模板＋说明＋脱敏案例。它帮你管住「一个人总控、多个长期 AI 助手」时最容易掉链子的几件事：活派出去有人收，说「做好了」有证据，写给新手的步骤照做就成，踩过的坑留下来给下一个人用。 借鉴的是方法思路；交付的是改写后的中文协作包，外加两支原创技能，并已在多个 Agent 工具里实际装上、按名调用过。借鉴来源见 NOTICE 和各技能末尾的来源／借鉴说明。 技能本身不带定时和叫醒（插件清单 hooks: {}）：到点查、记接…

项目简介：File browser, app data manager and cache cleaner for iOS 26 and 27 (worse_query, no jailbreak)

技术栈：Objective-C

核心介绍：and 27. “WQ” stands for worse_query, the sandbox technique it is built on. It reaches other apps’ and system containers with no jailbreak and no kernel exploit. Tested on iOS 26.7.1, 27.0.1, 27.2.1 and the 27.3 beta (confirmed by a user). containers there, but not System Data or the system groups, so WQ Files hides those two and the Cache Cleaner has no crash-log section. A full inode scan would take most of an hour on iOS 26, so App Data, Extension Data and…

项目简介：The open source ServerOS panel: one place to run the Linux servers you already own.

技术栈：PHP、agpl、control-panel、devops、laravel、open-source、self-hosted、server-management、serveros

核心介绍：ServerOS panel One place to run the Linux servers you already own. ServerOS runs the Linux servers you already own from one place. Connect a VPS from any provider with one command, then deploy apps, sites and databases, watch them,

项目数据：⭐ 747 Stars，🍴 59 Forks

项目简介：A live map beside the chat that shows what Claude Code is doing: tool calls, model steps, hooks, skills and subagents light up as they happen. Local only, in 7 languages.

技术栈：TypeScript、agent-observability、ai-agents、claude、claude-code、claude-code-plugin、developer-tools、svg-animation、typescript

核心介绍：A mod for Claude Code: a live map of the work, in a pane beside the chat. Every tool call, model step, hook and skill sends a pulse from the centre to its station, so you see memory, files, shell, web and agents at work instead of reading a list of commands.

项目数据：⭐ 729 Sta…

Frontier video generation models work in a VAE latent space that compresses video in both space and time. This compression made training tractable, but now limits high-resolution, controllable, and interactive generation — doubling resolution quadruples token count, conditions apply to groups of frames rather than individual ones, and each response must wait for the rest of its multi-frame block. We present PixVid, a simple yet effective method that lifts a pretrained latent video model to directly generate every frame in pixel space. PixVid keeps the diffusion transformer and replaces the…

Reference-frame stabilization registers every frame of a hovering-drone video onto one reference frame, keeping object positions comparable. The registration pipelines it relies on are selected on feature-matching benchmarks in which only the camera moves. Over a busy intersection, vehicles moving together supply mutually consistent correspondences that can draw the robust homography estimate toward their motion: consensus capture, which inlier statistics cannot flag. A two-part benchmark measures it and the object masking meant to prevent it. A static protocol scores ten pipelines under ex…

Machine-learning surrogates are increasingly used to accelerate computational fluid dynamics, yet progress is limited by the lack of benchmarks capturing realistic, time-dependent turbulent flows. We introduce TurtLES, a 14 TB dataset of 3D turbulent cylinder wakes in the shear-layer transition regime, computed via high-fidelity implicit large-eddy simulations. The dataset contains 400 trajectories of 400 time steps each, with 3-11 million points per frame, spanning variations in geometry, Reynolds number, and angle of attack. Unlike existing datasets, TurtLES combines (i) physically realis…

Existing adaptive-reasoning methods shorten the explicit reasoning a model writes, or skip it by switching modes; either way the answer is still written out token by token. Rather than deciding how much to think, we explore a different paradigm, zero-or-T thinking: a single model either answers from the implicit thinking inside one forward pass, generating no token, or performs full autoregressive reasoning at a cost of T tokens when that is worth paying. We realise it in a hybrid decision model (HDM), where one frozen backbone serves both modes and a lightweight self-routing block chooses …

Large language models (LLMs) increasingly mediate how people seek and consume information, including accounts that differ across languages and regions, especially on contested topics. Supporting diverse information seeking requires models to help users navigate competing narratives across languages and information streams. We define six capabilities necessary for diverse multilingual information seeking: three foundational capabilities for narrative inclusion, narrative exclusion, and missing narrative identification; and three higher-level capabilities for presenting diverse narratives, pr…

Medical imaging is essential for diagnosing disease, guiding treatment and monitoring patients over time. Although artificial intelligence has advanced image interpretation, existing medical models often address restricted tasks, while general-purpose proprietary models lack access to private clinical imaging required to achieve medical expertise. Developing models that can interpret complex examinations and communicate their findings in the context of clinical decisions remains a major challenge. Here, we show that Neuro-VL, a vision-language model trained on nearly one million neuroimagin…