2026年11月7日 技术热点总结

项目简介：YellowFruit third-party client

技术栈：TypeScript、android、drama、free、movie-web-app、reels、windows

核心介绍：Windows / macOS / Android 三端 · Fluent 设计 · 沉浸式竖屏播放 > ⚠️ 本项目仅供 学习与技术交流，不提供、不存储、不分发任何音视频内容。 > 所有内容均来自第三方站点，版权归原作者与上游站点所有。 > ⚖️ 许可证自 v1.1.0 版本起改为 PolyForm Noncommercial 1.0.0：禁止任何商业用途—— > 包括打包出售、接入付费/会员/广告变现、作为收费产品或服务的一部分分发。 > 个人学习、研究、自用与非营利使用不受限制，详见 LICENSE。 其余网站、网盘、QQ 群、公众号、第三方下载站分发的「HGhub」安装包一律非官方，作者不保证其安全性与完整性。 > 🚫 已发现的恶意修改版：hichuv/HGhub > 该仓库与本项目无关，是在旧版本源码基础上擅自加入卡密验证、VIP 付费解锁等变现功能的修改版， > 代码改动不透明、安全性无法验证。请勿下载、勿注册、勿付费；如已安装，建议立即卸载，改用官方 Release 包安装。
…

项目简介：Blazing-fast native notes app for macOS & Windows. Rust + GPUI, local-first, compatible with Feishu/Lark DocxXML.

技术栈：Rust

核心介绍：Quiet.md A fast, quiet, local notes app. Drawn natively with Rust + GPUI. Your notes stay on your computer, stored in the Feishu (Lark) DocxXML format. English · 简体中文 · 日本語 Website · Download · Performance · Features · Contributing > Quiet.md is in beta and changing quickly. If you hit a b

项目数据：⭐ 1,180 Stars，🍴 104 Forks

项目简介：A lightweight 2.5D character player for the web, with lip sync and webcam face tracking.

技术栈：TypeScript、avatar、character、face-tracking、lip-sync、pngtuber、vrm、vtuber、webgl2

核心介绍：A lightweight 2.5D character player for the web, with lip sync and webcam face tracking. A few lines of code put a character in a page, a stream overlay or an app. Comes with 21 featured characters in different styles, ready to use. English · 中文 · 日本語 (expressions, actions, microphone, camera; ?model=../characters// picks another character). /example/png.html and /example/vrm.html show a PNGTuber a…

技术栈：Python

核心介绍：本仓库收录两个独立版本：浏览器扩展版与批量版，分别保留各自的目录结构和原有代码。 https://linux.do/t/topic/3123997 （💰 免费激活 Google AI Pro 12 / 24 个月」| offer token 注入） （我本人大部分号是用的是注册机批量搞的，原理都一样，总共整了七个目前都活过了半个月，不过被这么大批量薅，估计也活不久了） > 在此提示，不要用大号！！！！ > PS：这里需要说明如果是已经给了折扣的情况，如下图，付款时有极大概率需要付款首月的youtube会员，小心别被反薅，正常显示的都是在下个月才收费 先把拓展切换到六个月，再重新进入页面，接着点击切换到年付 就会再次出现订购页面，这里会出现2028.5.5付款（示例日期），不要担心，这个其实还是免费的，到2028.5.5才会付款；同样的方法再次操作，切换下一个6个月，就能实现2年pro会员了 > 最后一定注意取消订阅，不要被反薅，因为附带有一个每月的youtube会员需要付费，不取消的话下个月就会被反薅

项目数据：⭐ 875 Stars，🍴 339 Forks

项目简介：Value your Steam stash in the terminal and print it as a share card: per game, per market, what sold and what sits. One Windows portable build, no account.

技术栈：Python、cli、inventory、price-check、python、report-card、steam、steam-market、valuation

核心介绍：paste a profile -> pick a card -> get a PNG worth posting or any custom app) from your terminal, prices them on Steam and on the big cash marketplaces, and prints the result as a shareable card: a report card, a game card, an inventory picture, a thermal receipt, a marketplace comparison, or a then-vs-now / A-vs-B diff. Every …

Offline retrieval-augmented generation (RAG) on edge devices is updated through signed knowledge packs that carry text and precomputed embedding vectors. A signature proves who published a pack, not that its vectors encode its text. We study this text-vector integrity gap. A verifier that recomputes vectors and accepts them within a cosine tolerance concedes a predictable share of queries to an attacker who controls the vectors: the hijack rate equals the distribution of angular retrieval margins at the tolerance (the margin law). Margins are small in five corpora and two embedding models, …

A deployable text-to-image diffusion model must hold several capabilities at once, such as instruction following, text rendering and human preference, and must sample in a few steps. Capabilities and few-step sampling are built in separate stages: a full-step model first consolidates capabilities through multi-teacher on-policy distillation, then serves as a fixed teacher for step distillation. Every full-step model we evaluate scores lower once compressed, and the second stage cannot begin until the first finishes. We introduce StepX-OPD, which merges the two stages: one student acquires e…

LLM-guided evolutionary search uses AI agents to improve programs for mathematical and computational problems by iteratively refining solutions and combining high-scoring methods. However, the process is often slow because the LLM generates a long textual chain of thought for each round and must repeat this generation many times to maintain population diversity. We present SwarmEvolve, a framework in which hundreds of low-cost agents per round work in swarms, reason in latent space without decoding to text, and share information by passing KV caches. Each swarm studies the problem from a di…

Large language models (LLMs) have shown increasing promise in research-level problem solving, yet a comprehensive collection of such problems is still missing, and evaluating them on research-level problems remains challenging. We introduce OpenFrontiers, a cross-disciplinary resource that combines a structured corpus of open research problems with a retrospective benchmark of recently resolved problems in mathematics. The corpus records problem statements, context, progress, and other information for 12,151 problems. The benchmark contains 200 problems, 110 resolved in 2026 and 90 still op…

Mathematical modeling turns an open-ended real-world problem into assumptions, a model, a solution, and a defended recommendation, and its answers matter only if they are both close to a credible reference and physically feasible; existing benchmarks either score solver fragments or grade reports with an LLM judge, leaving those numbers unchecked. We introduce TB-MathModel (the Terminal-style Mathematical Modeling Bench), the first benchmark to make complete contest-style modeling problems agentic, verifiable, and terminal at once, over 18 real CUMCM and MCM problems (2023–2025). Each task…

Linear attention compresses context into a fixed-size key-value state for linear-time training and constant-memory inference. Gated delta-rule models decay this state and erase its prediction for an incoming key before writing a new value; Gated DeltaNet-2 (GDN-2) generalizes this with channel-wise decay and decoupled, full-rank erase and write gates. However, we show that state updates are far less sensitive to erasure than to writes: key activation and unit normalization concentrate key energy onto a sparse subset of channels, whereas value writes scale all channels, so the erase gate can…