ICML2025-Conservative Offline Goal-Conditioned Reinforcement Learning

Lecture slides for the MARL book (www.marl-book.com)

[COLM 2025] Improving Fisher Information Estimation and Efficiency for LoRA-based LLM Unlearning

Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflo…

An end-to-end open ecosystem for robot learning

A PyTorch Library for Accelerating 3D Deep Learning Research

Code for "SceneSmith: Agentic Generation of Simulation-Ready Indoor Scenes"

Official implementation of Learning Athletic Humanoid Tennis Skills from Imperfect Human Motion Data

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

Fast and memory-efficient exact attention

HunyuanImage-3.0: A Powerful Native Multimodal Model for Image Generation

Official implementation of "HUSKY: Humanoid Skateboarding System via Physics-Aware Whole-Body Control"

Cosmos-Transfer1 is a world-to-world transfer model designed to bridge the perceptual divide between simulated and real-world environments.

Grounded SAM 2: Ground and Track Anything in Videos with Grounding DINO, Florence-2 and SAM 2

Cosmos-Transfer2.5, built on top of Cosmos-Predict2.5, produces high-quality world simulations conditioned on multiple spatial control inputs.

Decoupled Q-Chunking

Code for Tackling Long-Horizon Tasks with Model-based Offline Reinforcement Learning

Code for Scalable Offline Model-Based RL with Action chunking

[CoRL 2025] UniSkill: Imitating Human Videos via Cross-Embodiment Skill Representations

Official Release of Multistep Quasimetric Estimation (MQE)

Online Goal-Conditioned Reinforcement Learning in JAX. ICLR 2025 Spotlight.

Code for "Transitive RL: Value Learning via Divide and Conquer"

Flow Policy Gradients for Robot Control

Simulated experiments for "Real-Time Execution of Action Chunking Flow Policies".

A benchmark for offline goal-conditioned RL and offline RL

The official implementation of "Horizon Reduction Makes RL Scalable"

Cosmos Policy

Conditional Memory via Scalable Lookup: A New Axis of Sparsity for Large Language Models