A plug-and-play compiler that delivers free-lunch optimizations for both inference and training.

ATLAS by General Intelligence Capital — Self-improving AI trading agents using Karpathy-style autoresearch

AI agents running research on single-GPU nanochat training automatically

Scalable toolkit for efficient model reinforcement

Fast, Accurate, Lightweight Python library to make State of the Art Embedding

Async RL Training at Scale

AdaL vs Claude Code on Karpathy's Autoresearch benchmark

Training tiny models to prove hard theorems

My personal directory of skills, straight from my .claude directory.

Run OpenClaw more securely inside NVIDIA OpenShell with managed inference

AWS ECS/Fargate execution backend for Harbor benchmarks

Moloch is a minimalist, feature-rich Beamer theme for LaTeX presentations with a clean design and extensive customization options.

Comprehensive open-source library of AI research and engineering skills for any AI model. Package the skills and your claude code/codex/gemini agent will be an AI research agent with full horsepowe…

Public repository for Agent Skills

OpenClaw-RL: Train any agent simply by talking

The agent that grows with you

The open source coding agent.

Lightweight coding agent that runs in your terminal

Just like TextEdit on Mac but dedicated to Markdown.

Google Workspace CLI — one command-line tool for Drive, Gmail, Calendar, Sheets, Docs, Chat, Admin, and more. Dynamically built from Google Discovery Service. Includes AI agent skills.

Packages for Typst.

practice made claude perfect

Your own personal AI assistant. Any OS. Any Platform. The lobster way. 🦞

A lightweight alternative to OpenClaw that runs in containers for security. Connects to WhatsApp, Telegram, Slack, Discord, Gmail and other messaging apps,, has memory, scheduled jobs, and runs dir…

The official GitHub page for the survey paper "A Survey on Evaluation of Large Language Models".

A lightweight, AI-native training framework for large language models. Designed for fast iteration, reproducible experiments, and modular configuration across SFT, RLVR, and evaluation workflows.