A neural network framework for instance segmentation of large, complex 3D shapes, particularly in brain tissue electron microscopy datasets.

A Node.js library that provides plug‑in “skills” for AI coding assistants (Claude Code, Copilot, etc.) to generate Selenium, Playwright, Cypress, and other test‑automation code and run it on the TestMu AI cloud platform.

A data and evaluation toolkit for non-intrusive load monitoring (NILM) that provides tools for dataset conversion, preprocessing, and performance metrics.

A collection of reproducible “recipes” (example projects) for the VeRL library, which provides reinforcement‑learning tools for large language models. Each recipe pins a specific VeRL version via a `REQUIRED_VERL.txt` file, and a helper script (`install_verl.sh`) can automatically install the matching library and run the example. The repo includes demos such as RL‑based tool use, LangGraph math agents, Single‑stream Policy Optimization, and more.

A persistent knowledge base and shared memory layer for AI agents that connects to various data sources and records session transcripts to maintain context across runs.

A GPU-accelerated and differentiable Python library for high-performance simulation of quantum systems, designed for large-scale problems and gradient-based parameter estimation.

A PyTorch implementation of the Dreamer 4 world model designed for multi-task continuous control, featuring a causal tokenizer and an interactive dynamics model.

An AI-powered computer automation tool that uses LLMs to self-drive a PC by simulating keyboard and mouse inputs based on visual feedback.

MAGI-1 is an autoregressive video generation model that produces high-fidelity videos chunk-by-chunk to ensure temporal consistency and physical accuracy.

KeyBERT is a lightweight Python library that extracts keywords/keyphrases from any text by comparing BERT‑based document embeddings with embeddings of candidate n‑grams. It works out‑of‑the‑box with a single `pip install` and three lines of code, supports multiple embedding back‑ends, and offers diversity controls (max‑sum, MMR) and an optional LLM‑based mode.

A cost-efficient routing system that replaces expensive LLM classification calls with lightweight traditional ML surrogates for easy inputs, maintaining teacher-level parity.

A personal AI agent that builds an evolving Personal Model of the user to help design and manage long-term life and work paths.

An open-source tool for hardening GenAI applications by simulating dynamic LLM-based attacks to test and refine system prompt security.

A collection of system prompts and documentation from Xcode 27 beta, detailing Apple's internal instructions and strategies for its AI-assisted coding tools.

A high-performance GPU-accelerated molecular dynamics package that supports empirical and neuroevolution machine-learned potentials for materials simulation.

A tool that installs and manages official AI skills provided by libraries, ensuring AI agents use the most current and correct code patterns instead of outdated training data.

Dockerfiles that provide a web-accessible Ubuntu Desktop environment pre-installed with ROS 2, allowing users to run robotics software via a browser.

A comprehensive collection of 32 design patterns and code examples for building reliable, scalable, and secure Generative AI applications.

A portable mixed-precision linear algebra library providing high-performance SIMD kernels for multiple CPU architectures to prevent numerical overflow and reduce binary size.

Conversation Stenography is an open‑source Go CLI that encrypts a secret message with AES‑SIV, encodes the ciphertext into token choices of a locally‑run LLM (e.g., Llama 3.2 or GPT‑2), and outputs natural‑looking cover text. Users copy that text into any chat app; the recipient runs the same tool to decode and decrypt the hidden message. The system is fully offline, uses a shared secret phrase (PBKDF2‑derived key), and chains messages for integrity. It includes a setup wizard, simulation mode, and several CLI commands, but is a proof‑of‑concept and vulnerable to existing steganography detection methods.

A clock-driven Python simulator for spiking neural networks designed to help scientists efficiently model biological neural activity.

A composable multi-agent shell that integrates natural language and bash commands into a single interface for seamless AI-driven command-line experiences.

A CLI tool for mobile developers to generate app icons and Google Play feature graphics using OpenAI and Google Gemini image models.

A simulation framework built on NVIDIA Isaac Sim for multirotor vehicles, providing integration with PX4 and ArduPilot flight control stacks.

A structured skill framework for Claude Code that adds 131 professional skills for deployment, security, and development via an MCP skill loader.

Ark is a Kubernetes-native framework for building and hosting portable, provider-agnostic AI agents using a declarative approach to avoid vendor lock-in.

An open agent operating system that provides a composable environment for AI agents to run, manage, and extend their own capabilities via modular capsules.

A minimalistic and portable C-based computer vision library providing state-of-the-art algorithms for face, object, and text detection.

A massive, open-source Chinese corpus project aiming to collect and provide 253TB of diverse internet text and multimodal data for training AI models.

A Python library for Chinese NLP preprocessing and parsing, providing tools for data cleaning, entity extraction, and automatic LLM evaluation.

An open-source Python library that provides model compression and quantization techniques for PyTorch, TensorFlow, and JAX models to improve inference efficiency.

ColPali is a research library for visual document retrieval. It encodes document images with vision‑language models (e.g., PaliGemma, Qwen‑VL) into multi‑vector embeddings and scores them against query embeddings using a ColBERT‑style late‑interaction method. The original `colpali-engine` package (now deprecated) provides model classes, processors, optional fused kernels, fast‑Plaid indexing, token‑pooling, and interpretability tools. New projects should use Sentence‑Transformers v6’s `MultiVectorEncoder`, which integrates ColPali‑style models.

A general-purpose probabilistic programming package for Bayesian and likelihood-based inference implemented in Julia.

A multi-agent system that automates the generation of high-quality, editable PPTs by coordinating specialized agents for outlining, parallel research, and quality checking.

A collection of standardized experiments and benchmarks for evaluating the core capabilities of reinforcement learning agents to facilitate reproducible research.

A general-purpose AI agent system that executes tasks in a secure Docker sandbox, supporting real-time browser and code execution.

A comprehensive Python library of 233+ meta-heuristic algorithms for solving complex global optimization problems and tuning AI/ML parameters.

A modular library of samplers for JAX that enables composable Bayesian inference on CPU, GPU, and TPU without requiring a full probabilistic programming language.

A local development gateway that lets you run frontier LLMs in your code editor using free browser sessions or local providers, bypassing expensive API keys.

A Kubernetes operator suite that automates the deployment, scaling, and management of LLM inference, fine-tuning, and RAG engines.

A C++17 library for simplifying CUDA kernel benchmarking, offering parameter sweeps and detailed throughput and bandwidth measurements.

An open-source Agent OS for intelligent hardware that transforms smart home devices into proactive AI agents using a brain-skill-memory architecture.

OpenRAVE is a software framework for robotics research and development focused on kinematics and dynamics.

An open-source neural network framework written in C/C++ and CUDA that powers the YOLO real-time object detection system.

A high-performance infrastructure for multi-dimensional sparse attention, implementing Neighborhood Attention to reduce the computational cost of self-attention for 2D and 3D data.

An open-source API proxy for Midjourney that allows developers to integrate Midjourney's image generation and face-swapping capabilities into their own applications via a programmable interface.

A collection of high-quality, open-source German voice datasets used to train offline, license-free text-to-speech (TTS) AI models.

An open-source e-commerce system based on Laravel 13 that integrates AI models and Model Context Protocol (MCP) for intelligent automation.