Neural Harmonic Textures for High-Quality Primitive Based Neural Reconstruction

A toolbox for real-to-sim reconstruction and robotic simulation

Open-source code of the paper: Real-to-Sim Robot Policy Evaluation with Gaussian Splatting Simulation of Soft-Body Interactions.

A unified, high-performance framework for training LLMs, VLMs, diffusion, and embodied models on NVIDIA GPUs and Kunlun XPUs.

GTSAM is a library of C++ classes that implement smoothing and mapping (SAM) in robotics and vision, using factor graphs and Bayes networks as the underlying computing paradigm rather than sparse m…

[RAL-25] An online open-vocabulary mapping system that enables natural language querying to navigate dynamic scenes, with ROS support.

🐥A PyTorch implementation of OpenAI's finetuned transformer language model with a script to import the weights pre-trained by OpenAI

Open-source HandUMI software for synchronized bimanual data collection and retargeting to any bimanual robot. Includes calibration, QA, replay, and teleoperation.

Grabette: robotic data recording device

[ICLR 2026] π^3: Permutation-Equivariant Visual Geometry Learning

Dimensional is the agentic operating system for physical space. Command humanoids, quadrupeds, drones, and other hardware platforms in natural language and build multi-agent systems that work seaml…

A feed-forward 3D foundation model for reconstructing scenes from streaming data

My solution for the LeHome challenge (1st place online, 2nd place in real world round)

Unsloth is a local UI for training and running Kimi K3, Gemma 4, Qwen3.6, DeepSeek-V4, GLM and other models.

Official repository of B-spline Policy: Accelerating Manipulation Policies via B-spline Action Representations.

Code for Xiaomi-Robotics-1

RoboDojo Official Repo

Source codes for the paper "RoboWits: Unexpected Challenges for Robotic Creative Problem Solving"

WARP-RM: A Warp-Augmented Relative Progress Reward Model for Data Curation

MME-VLA-Suite

A Python port of Pi’s minimalist coding agent.

Improved HaWoR fork: temporally stable, lower-jitter egocentric hand pose via zero-lag smoothing + bone canonicalization (no retraining)

Efficient Triton Kernels for LLM Training

[ICLR 2025] OpenVid-1M: A Large-Scale High-Quality Dataset for Text-to-video Generation

LightGlue: Local Feature Matching at Light Speed (ICCV 2023)

Video-Action Models for Generalizable Robot Control Beyond VLAs