Official github repo for CVPR 2026 paper DreamSAC: Learning Hamiltonian World Models via Symmetry Exploration

[CVPR 2026 Oral, Best Paper Finalist] SeaCache: Spectral-Evolution-Aware Cache for Accelerating Diffusion Models

nanobind: tiny and efficient C++/Python bindings

1 million FPS multi-agent driving simulator

[NeurIPS 2025] RAD: Training an End-to-End Driving Policy via Large-Scale 3DGS-based Reinforcement Learning

Continued development of puffer/drive

58 implementations of synthetic learning problems from Jürgen Schmidhuber's papers (1989-2025). Pure numpy, laptop-runnable, paper-comparison metrics per stub. Algorithmic-lineage companion to hint…

Use Garry Tan's exact Claude Code setup: 23 opinionated tools that serve as CEO, Designer, Eng Manager, Release Manager, Doc Engineer, and QA

AI agent skill that researches any topic across Reddit, X, YouTube, HN, Polymarket, and the web - then synthesizes a grounded summary

AI coding assistant skill (Claude Code, Codex, OpenCode, Cursor, Gemini CLI, and more). Turn any folder of code, SQL schemas, R scripts, shell scripts, docs, papers, images, or videos into a querya…

A document containg an explicit explanation of the process used by Gaussian Splatting to render images to the screen

A complete academic research Skill suite. Supports Claude Code, ChatGPT / Codex CLI, and Gemini CLI.

An open-source, GPU-accelerated physics simulation engine built upon NVIDIA Warp, specifically targeting roboticists and simulation researchers.

The official repository of Terra Nova, a comprehensive challenge environment for reinforcement learning research inspired by Civilization V.

A framework for creating rich, 3D, Minecraft-like single and multi-agent environments for AI research. (Accepted at ICML 2025).

Minimal and annotated implementations of key ideas from modern deep learning research.

16-bit CPU emulator from scratch in pure C

A platform for reproducible world model research and evaluation

Experiment with coding agent to train a decoding transformer to perform addition

C++ implementation of Andrej Karpathy's microgpt

Official Implementation of the ICCV 2023 paper: Perpetual Humanoid Control for Real-time Simulated Avatars

A curated list of awesome exploration RL resources (continually updated)

A minimalist, high-performance Astro theme for photographers and artists.

Puzzles for learning Triton

Softlearning is a reinforcement learning framework for training maximum entropy policies in continuous domains. Includes the official implementation of the Soft Actor-Critic algorithm.

CUDA-L1: Improving CUDA Optimization via Contrastive Reinforcement Learning

KernelBench: Can LLMs Write GPU Kernels? - Benchmark + Toolkit with Torch -> CUDA (+ more DSLs)

Fast and memory-efficient exact attention

Tile primitives for speedy kernels