Build AI text detectors whose decisions are mechanistically grounded and human-auditable (EMNLP 2026)
Search your computer by what's in your files: photos, PDFs, documents, recordings, videos and code. On-device on Mac, Windows and Linux, built on Google DeepMind's EmbeddingGemma 2.
Multi-Joint dynamics with Contact. A general purpose physics simulator.
GDM Science Skills to speed up agentic scientific workflows with better grounding and higher token efficiency. Integrate insights from AlphaGenome, AFDB, UniProt and 30+ other databases and tools.
Open-source implementation of AlphaEvolve
A collection of high-quality models for the MuJoCo physics engine, curated by Google DeepMind.
AlphaFold 3 inference pipeline.
Better CarPlay for MHI2Q! Cluster Map, HUD Projection, Spatial Audio and more
—
A library for generative social simulation
Interview-driven AP242 PMI authoring for STEP parts in Rust and Python
IP-KVM on the ESP32-P4: HDMI capture, USB keyboard and mouse, and a browser console over HTTPS
Open and inexpensive DIY IP-KVM based on Raspberry Pi
Gemma open-weight LLM library, from Google DeepMind
OpenSpiel is a collection of environments and algorithms for research in general reinforcement learning and search/planning in games.
GPU-optimized version of the MuJoCo physics simulator, designed for NVIDIA hardware.
A collection of formalized statements of conjectures in Lean.
A next-generation automatic algorithm design platform, making automated algorithm design more accessible and easier to use
基于 Google DeepMind 的 EmbeddingGemma 2 的本地多模态文件语义搜索器:用一句话描述文件的内容,在你选定的文件夹中找到文档、照片、截图、录音和视频,也可以以图搜图。支持 Windows、Android。
An open-source library for sim and real robot demos on top of MuJoCo.
AlphaProtein Novo
Google DeepMind's software stack for physics-based simulation and Reinforcement Learning environments, using MuJoCo.
Real-time behaviour synthesis with MuJoCo, using Predictive Control
Resident Evil project to share classic survival horror with friends through early experimental split-screen support.
Research code accompanying AlphaGenome
Tracking Any Point (TAP)
😎 Awesome list about Gemma, Google DeepMind's family of lightweights, state-of-the-art open models.
Amplio: A Lightweight Agent Harness for Robust and Long-Horizon Runs
Optax is a gradient processing and optimization library for JAX.
µStreamer - Lightweight and fast MJPEG-HTTP streamer
RL research on Android devices.
A suite of test scenarios for multi-agent reinforcement learning.
Some Cryptids and Lean proofs
This API provides programmatic access to the AlphaGenome model developed by Google DeepMind.
TORAX: Tokamak transport simulation in JAX
—
bsuite is a collection of carefully-designed experiments that investigate core capabilities of a reinforcement learning (RL) agent
—
Remote IPMI exporter for Prometheus
—
Minimal and scalable research codebase in JAX, designed for rapid iteration on frontier research in LLM and other autoregressive models.
—
Custom VA-API video driver for AMD BC-250, using Vulkan RDNA 2 compute shaders to enable H.264 hardware-accelerated encoding.
Free, open-source AI clip generator and Opus Clip alternative. Turns podcasts, vlogs, tutorials and stream VODs into captioned YouTube Shorts, Reels and TikToks on your own PC: local Whisper, LLM ranking, speaker-tracking crop, YouTube upload. Windows, Mac, Linux. No subscription, no watermark.
A library of reinforcement learning components and agents
JAX-based neural network library
Monte Carlo tree search in JAX
Open and cost-effective "KVM-over-IP". BliKVM comes in 4 different models, v1 CM4, v2 PCIe, v3 HAT and v4 Allwinner, based on Raspberry Pi and Allwinner SoC.
Code repo for xCAT core packages
Spatial audio for any headphones on Mac: 7.1 surround with head tracking for every app, Netflix in 5.1 in Safari, AI upmix for stereo, and a live 3D view of the sound.
SynthID Bio is a family of methods developed by Google DeepMind for embedding highly detectable yet function-preserving watermarks directly into AI-generated biological sequences and structures.
cnpm: npm client for China mirror of npm
tree is a library for working with nested data structures
—
PettingZoo and Gymnasium bindings for popular reinforcement learning environments outside of Farama
A platform for managing machine learning experiments
An OSGi-like C++ dynamic module system and service registry
SymmetricDS is database replication and file synchronization software that is platform independent, web enabled, and database agnostic. It is designed to make bi-directional data replication fast, easy, and resilient. It scales to a large number of nodes and works in near real-time across WAN and LAN networks.
An implementation of the Fermionic Neural Network for ab-initio electronic structure calculations
—