2026年10月7日 技术热点总结

项目简介：An open-source, clean-room reimplementation of Adobe Premiere Pro built in pure Rust.

技术栈：Rust、adobe、art、rust、video、video-editing、video-editor、video-processing

核心介绍：FilmCraft Video editing, color and sound; an open-source, clean-room reimplementation of Adobe Premiere Pro, rebuilt in pure Rust. An open-source, clean-room take on the Adobe Premiere Pro workflow: native on macOS, Windows and Linux, and in the browser via WebAssembly. By the ArtCraft team. FilmCraft on getartcraft.com · ArtCraft · All Crafting Apps

项目数据：⭐ 1,596 Stars，🍴 442 Forks

技术栈：Lean

核心介绍：This repository contains mathematical manuscripts and supporting proof artifacts produced by an internal OpenAI model. As part of model development, we evaluate our models on open research problems. We expanded these evaluations after performance on our existing mathematical evaluations saturated. Some outputs build upon earlier results produced by the models.

项目数据：⭐ 1,279 Stars，🍴 96 Forks

技术栈：C++

核心介绍：THIS PROJECT IS NOT RELATED TO SHADPS4. ALL QUESTIONS RELATED TO THIS PROJECT SHOULD BE SENT TO THE DISCORD SERVER https://discord.gg/KYZRKk9CB, NOT TO THE SHADPS4 SERVER. bbport is the counterpart of Wine + DXVK for a single game: *Bloodborne* for PlayStation 4 (CUSA03173, game version 1.09) on an x86-64 Linux PC. The game’s original executable runs this one game replaces the PS4 system libraries; shadPS4 and heavily extended for this game, including

关键特性：Native execution. The eboot is converted offline into a flat memory image; PS4 libc and；**PC memory model…

项目简介：An open-source, clean-room reimplementation of Adobe Lightroom in pure Rust.

技术栈：Rust、art、lightroom、photography、raw、raw-editor、raw-processing、rust

核心介绍：LightCraft Your photos. Your pixels. Your machine. Photo library and raw development; an open-source, clean-room reimplementation of Adobe Lightroom, rebuilt in pure Rust. Native on macOS, Windows and Linux. In the browser via WebAssembly. Drivable end to end by AI agents over MCP. LightCraft on getartcraft.com · ArtCraft · All Crafting Apps

项目数据：⭐ 1,213 Stars，🍴 347 Forks

项目简介：An open-source, clean-room reimplementation of Adobe Acrobat built in pure Rust

技术栈：Rust、acrobat、acrobat-full、acrobat-pro、adobe、adobe-acrobat、document-processing、documents、pdf

核心介绍：PrintCraft The PDF workbench; an open-source, clean-room reimplementation of Adobe Acrobat, rebuilt in pure Rust. Read, organize, combine, split and secure PDFs in a fast, native app, written in Rust from the ground up. macOS · Windows · Linux · FreeBSD · the web PrintCraft on getartcraft.com · ArtCraft · All Crafting Apps The PrintCraft Showcase, a 13-page specimen PDF, open with the All…

项目简介：An open-source, clean-room reimplementation of Adobe Illustrator, built in pure Rust.

技术栈：Rust、adobe-illustrator、art、illustrator、illustrator-2023、illustrator-extensions、illustrator-free、illustrator-full、rust

核心介绍：VectorCraft Vector illustration; an open-source, clean-room reimplementation of Adobe Illustrator, rebuilt in pure Rust. A fast, open-source, clean-room take on the Adobe Illustrator workflow. It runs natively on macOS, Windows, Linux and FreeBSD, and in the browser via WebAssembly. Built by the ArtCraft team. VectorCraft on getartcraft.com · ArtCraft · All…

项目简介：明日方舟「卫戍协议：盟约」非官方同人复刻：浏览器自走棋塔防，单人或 1–4 人联机合作（非商业）

技术栈：JavaScript、arknights、auto-chess、fan-game、nodejs、tower-defense、websocket

核心介绍：《明日方舟》季节性自走棋塔防玩法「卫戍协议：盟约」的非官方同人复刻：浏览器即开即玩，单人或 1–4 人联机合作。 > – 本项目是玩家自制的非官方同人作品，与上海鹰角网络科技有限公司（Hypergryph）、Yostar 及其关联方没有任何关系，未获其授权或认可。 > – 《明日方舟》及「卫戍协议」相关的名称、角色、美术、音乐、音效、文本与数据等素材，版权归原权利人所有。这些素材不适用本项目的 GPL-3.0 许可证；GPL 只覆盖本项目自己编写的代码。 > – 仅供学习交流与个人非商业使用。严禁任何形式的盈利，包括但不限于：售卖本项目或整合包、付费下载或付费分发、收费服务器或收费代开、广告 / 打赏 / 会员等变现方式，以及其他任何商业用途。 > – 仓库源码不包含游戏的美术与音频素材（只有由官方数据表生成的数据和几张游戏截图，同样不适用 GPL）；Releases 中的整合包为了方便玩家附带了素材，下载即视为同意本声明。请勿将素材用于本项目…

Does conversational memory need LLM-extracted facts, or is selecting the right raw turns enough? Published results disagree. Extraction-based systems report gains from distilled facts. Recent studies find raw history with good ranking does as well, but disagree about whether ranking matters. We ran a pre-registered study on held-out LoCoMo conversations and LongMemEval. At a tight budget on LoCoMo, raw turns selected by a single call to Jev, a typed decision model, are non-inferior to an LLM-extraction memory (one-sided 95% bound -3.0 points against a -5-point margin). Blind human grading n…

Numerical Linear Algebra (NLA) has consistently played a vital role in advancing science by providing tools to solve fundamental problems encountered in scientific and engineering applications. Over the decades, it has continually evolved to meet the demands driven by successive waves of scientific discovery. For instance, during the 1950s and 1960s, substantial efforts were devoted to developing methods for solving eigenvalue problems that emerged from the rapidly growing field of aerodynamics. This led to the discovery of the LR and QR algorithms. Later the attention turned to the solutio…

Routine complete blood counts (CBCs) could yield new biomarkers, but the private records needed to evaluate candidates cannot be shared with frontier language model agents that excel at discovery. We distilled the evidence held in the Clalit Health Services panel of over 5.4 million patients into a released scoring tool: for each of 13 immune-mediated diseases, a graph attention network was trained inside the data boundary to predict the case-control AUC of candidate CBC expressions, and only the trained weights were released. The tool grounds an agent's propose-score-refine loop in real-wo…

Tabular foundation models perform in-context learning (ICL) by conditioning predictions on labeled training examples provided as context. Unlike traditional models that separate training from inference, these models must process all training examples in every forward pass, making each prediction expensive. Restricting the number of training examples reduces this cost but substantially degrades performance. Instead of discarding context, we propose activation alignment, a method that leverages the full context to teach a model how to behave when seeing only a subset. This is achieved by trai…

A language model can give a correct answer more probability than any single incorrect answer and still usually sample an incorrect one, because the incorrect answers together hold more probability. The power distribution raises each complete answer's probability to a power above one and renormalizes, shifting probability toward answers the model finds most likely (sharpening). Sampling from it improves reasoning without changing parameters, but needs many scored candidates per query. We show that a model can instead be trained to produce such answers in one generation. On-policy power disti…

A typed decision model answers a fixed question about an input by returning a probability for each of several caller-defined options. Each option carries a short label and a written definition, which is where a developer states the rule the model should apply. Jev introduced this interface for routing, moderation and triage, open implementations followed, and the same operation occurs whenever a language model is used as a classifier by scoring label strings. We study the open implementations, whose weights we can inspect and patch, and ask whether the probability follows the definitions or…