An open-source adversary emulation framework for red-teaming AI agents and MCP servers, providing automated vulnerability scanning based on OWASP standards.

A collection of Topological Data Analysis (TDA) Python libraries designed to make topological data tools accessible to non-topologists.

ONE is a high-performance on-device neural network inference framework that enables AI models from TensorFlow and PyTorch to run efficiently across various hardware accelerators.

A comprehensive evaluation library for NLP and LLM models to test for robustness, bias, fairness, and safety.

An R package that brings PyTorch functionality to the R ecosystem, enabling tensor operations and automatic differentiation for machine learning.

A PyTorch-based Python library designed for deep learning research in hydrology, allowing users to train neural networks via configuration files.

A framework that enables AI agents to generate interactive React-based HTML artifacts instead of plain Markdown, allowing for complex layouts, dashboards, and stateful interfaces.

A library providing a programming model for writing performance-critical machine learning kernels that are portable across different GPU and CPU architectures.

A Java thread dump and heap analyzer that provides a GUI and an MCP server to help developers diagnose deadlocks, performance bottlenecks, and virtual thread issues.

OSCAR is a 2-bit KV cache quantization method that uses offline spectral covariance-aware rotations to reduce memory usage by 8x while maintaining near-BF16 accuracy.

JADE is a safety evaluation platform that uses linguistic mutation to generate high-risk datasets for testing the safety guardrails and alignment of LLMs and AI agents.

A symbolic modeling package for the high-performance simulation and analysis of chemical reaction networks and dynamical systems.

A local orchestration tool that uses multiple AI subscriptions to provide parallel, multi-vendor code reviews to catch bugs and biases that a single AI model would miss.

FlagScale is a unified toolkit for the FlagOS ecosystem that provides a single interface for training, reinforcement learning, and inference of large models across diverse AI hardware chips.

An engine for ML and data tracking, visualization, and explainability that integrates with major deep learning frameworks to log metrics, inputs, and artifacts.

An LLM runtime that enables running massive Mixture-of-Experts (MoE) models on consumer NVIDIA GPUs by managing expert residency between VRAM and CPU RAM.

A collection of reference implementations of DeepMind's DQN-based reinforcement learning agents, designed for research reproducibility on Atari games.

An AI-powered SQL client and database administration tool supporting multiple popular databases for table design, querying, and synchronization.

A self-attention-based framework for multivariate time series imputation that solves the speed and memory issues of RNN-based models.

A library of traditional machine learning algorithms for Elixir, providing tools for classification, regression, clustering, and preprocessing built on top of Nx.

A PyTorch-based module for implementing and testing Echo State Networks and Reservoir Computing, designed for research purposes and integration into deep architectures.

A framework for scaling reasoning-capable models through a self-improvement loop of synthetic data generation and automated verification.

A Python library for generating scalable synthetic data within Databricks using Spark, designed for testing, benchmarking, and demos.

A PyTorch-based training framework for Stockfish's NNUE evaluation networks, enabling GPU-accelerated training and automated performance evaluation.

A framework for constrained nonlinear optimization designed for scientific machine learning, uncertainty quantification, and AI, offering fine-grained control over solver workflows.

A framework for evaluating and comparing open-source AutoML systems using standardized environments and curated datasets to ensure reproducible research.

A set of plugins for radare2 that integrates LLMs to assist in reverse engineering tasks such as function explanation, vulnerability discovery, and augmented decompilation.

A Clojure library for fast tensor and neural network computations that leverages optimized native libraries for CPU and GPU acceleration.

An integration library that allows JAX users to call custom Triton kernels for high-performance GPU operations within JAX programs.

An enterprise-grade multimodal AI middleware that provides a unified API layer for managing multiple model providers, RAG, and agent runtimes with centralized routing and governance.

A PyTorch library for efficiently computing the Hessian eigendecomposition, trace, and spectral density of neural networks without the quadratic memory bottleneck.

A toolchain for deploying optimized AI algorithms on low-cost embedded devices and microcontrollers, supporting computer vision, anomaly detection, and scenario-specific tasks.

A web-based GUI for Ollama and other local LLM providers that allows users to chat with self-hosted models via a privacy-focused interface.

An AutoML technology for tabular data that synthesizes high-quality machine learning pipelines by learning from a corpus of human-written solutions.

A Python-based framework that allows developers to author, debug, and optimize ONNX models using a natural Python syntax instead of manual graph construction.

A zero-dependency local dashboard for OpenClaw AI agents that provides real-time monitoring of costs, system health, and agent activity.

A Julia package for data-driven structural estimation and identification of differential equations, allowing users to automatically discover mathematical models from data.

A high-performance MCP management system built with Rust that orchestrates Model Context Protocol services for AI agents and applications.

An autonomous, AI-driven penetration testing framework that orchestrates over 72 security tools to automate reconnaissance, exploitation, and reporting.

A scalable PyTorch reinforcement learning framework accelerated by Lightning Fabric that decouples RL algorithms from their environments for greater flexibility.

A PyTorch-native toolbox for discrete and continuous wavelet transforms that provides GPU acceleration and gradient support for signal and image processing.

A toolkit for developing and deploying TFLite models to mobile devices, providing tools for metadata management, automatic API generation, and optimized task libraries.

A command-line tool to execute .prompt files, allowing users to treat LLM prompts as versionable artifacts and automate AI workflows via the shell.

A PyTorch-based parallel training system designed to scale deep learning recommendation models up to 100 trillion parameters on commodity hardware.

A Python package for automated LLM compression that uses advanced quantization techniques like QEP and AutoBit to reduce model size and VRAM usage.

A library for large-scale and reproducible hyperparameter optimization (HPO) that supports distributed execution and state-of-the-art tuning methods.

An AI-native SDLC workflow for Java Enterprise applications that provides reusable skills, agents, and commands to standardize analysis, design, and implementation.

CRAB is a Python-centric framework for building and evaluating multimodal embodied AI agents across diverse environments like Docker, VMs, and physical machines.