Starred repositories
Reproduce ABC-130K (abc.bot) on a desktop bimanual station — plus pi0 / pi0.5 and your own policies. Full stack from teleop collection to real-robot rollout, on off-the-shelf i2rt YAM hardware.
MANO hand models in MJCF format for MuJoCo simulation.
QQTang是一个面向经典游戏《QQ堂》的非商业怀旧与技术研究项目，目标是在保留原版客户端体验的基础上，通过逆向工程与协议分析逐步恢复登录、大厅、房间、竞技、探险、商城、BOSS、装扮与多人联机等核心功能
Hypersim: A Photorealistic Synthetic Dataset for Holistic Indoor Scene Understanding
RL-training knowledge-work agents on APEX-Agents data with SkyRL + Harbor
Path tracing renderer and utilities for three.js built on top of three-mesh-bvh.
A Modern Python and C++20 Library of Unified Incremental Potential Contact.
Community plugin to control Blender 3D with any LLM of your choice. Not affiliated with the official Blender Foundation.
Isaac Lab API, powered by MuJoCo-Warp, for RL and robotics research
Routines in text format for importing into HeavySet
converting indoor point cloud scans into structured data
(ICLR2026) ReconViaGen: Towards Accurate Multi-view 3D Object Reconstruction via Generation
Native and Compact Structured Latents for 3D Generation
OpenTopos — unified, code-driven 3D object generation: static + articulated + texture (incl. UV atlas), multi-agent + Blender
RoboVerse: Towards a Unified Platform, Dataset and Benchmark for Scalable and Generalizable Robot Learning
A lightweight sandboxing tool for enforcing filesystem and network restrictions on arbitrary processes at the OS level, without requiring a container.
A research microVM for macOS on Apple Silicon (Hypervisor.framework) — architecturally modeled on AWS Firecracker, not a port. Boots Linux, virtio, SMP, snapshot/restore.
Rebuild the object in a reference image as a code-only, procedural, quality-gated, animation-ready Three.js model. Token-efficient image-to-3D.
Codex plugin that turns attached object images into code-only, animation-ready procedural Three.js models.
Lightweight coding agent that runs in your terminal
FlashInfer: Kernel Library for LLM Serving
[CVPR 2026] Scaling Spatial Intelligence with Multimodal Foundation Models