ws2644
Follow
Stars
A studio for designing and shipping shadcn-style components in Expo/React Native with Storybook-backed visual regression.
A manifesto and playbook for AI-native software engineering in the LLM era / AI-Native的软件工程宣言
HunyuanVideo-1.5: A leading lightweight video generation model
NEO Series: Native Vision-Language Models from First Principles
SAG - SQL驱动的RAG引擎 · 查询时自动构建知识图谱 | SQL-Driven RAG Engine · Automatically Build Knowledge Graph During Querying
A reading list for trustworthy audio large language models.
GigaBrain-0: A World Model-Powered Vision-Language-Action Model
A lightweight browser-to-NAS pipeline for capturing and downloading web videos. It integrates a Chrome Extension with a NAS-hosted Docker backend (FastAPI, workers, FFmpeg) to automatically detect,…
PageEyes Agent 是一个轻量级 UI Agent，通过自然语言指令驱动，无需编写脚本既可实现Web、Android平台的UI自动化任务。
High-quality and compute-verified reproductions of cutting-edge AI papers.
[NeurIPS'2025] Official repository for "LiveStar: Live Streaming Assistant for Real-World Online Video Understanding"
The first Interleaved framework for textual reasoning within the visual generation process
The Intelligent GUI Agent for Mobile Phones
AipexBase is an AI-native BaaS platform. You only need to develop the frontend with vibe coding tools, and leave the backend to AipexBase!
PromptEnhancer is a prompt-rewriting tool, refining prompts into clearer, structured versions for better image generation.
🔥[ICML 2024, Official Code] First work to propose a solution to the long-tail problem in IAA. 首篇针对IAA中的长尾问题提出解决方案的工作
react markdown typing animation component
This is the official implementation for the paper "SNR-aware low-light image enhancement" in CVPR2022
UCCL is an efficient communication library for GPUs, covering collectives, P2P (e.g., KV cache transfer, RL weight transfer), and EP (e.g., GPU-driven)
[NeurIPS'25] Official repository of Concerto: Joint 2D-3D Self-Supervised Learning Emerges Spatial Representations
Official codebase for "Brain Harmony: A Multimodal Foundation Model Unifying Morphology and Function into 1D Tokens" (NeruIPS 2025).
对传统rag的改造，让大模型像专业研究员一样读懂厚重的基础设施公募Reits招募说明书，精准定位证据并复核答案。