---
orphan: true
---

# RoboVerse

**RoboVerse** is the dataset, benchmark, task-pack, and learning layer that sits on
top of the [MetaSim](https://roboverse.wiki/metasim/) simulator framework. This
section brings together the task inventory, asset descriptions, benchmark protocol,
and the imitation- and reinforcement-learning workflows.

---

## Quick Navigation

::::{grid} 2
:gutter: 3

:::{grid-item-card} Tasks Overview
:link: /dataset_benchmark/tasks/overview
:link-type: doc

Browse the full task inventory across every benchmark namespace.
:::

:::{grid-item-card} Robots, Objects, Scenes
:link: /dataset_benchmark/dataset/robots
:link-type: doc

Robot configurations, object assets, and scene definitions used by tasks.
:::

:::{grid-item-card} Imitation Learning
:link: /roboverse_learn/imitation_learning/diffusion_policy
:link-type: doc

Diffusion Policy, ACT, OpenVLA, SmolVLA, RDT, Octo and how to add your own.
:::

:::{grid-item-card} Reinforcement Learning
:link: /roboverse_learn/reinforcement_learning/ppo
:link-type: doc

PPO, FastTD3, SAC, TD3, SkillBlender, humanoid locomotion stacks.
:::

:::{grid-item-card} Benchmark Protocol & Results
:link: /dataset_benchmark/benchmark/overview
:link-type: doc

Standard evaluation protocols and the latest benchmark results.
:::

:::{grid-item-card} MetaSim docs ↗
:link: /metasim/
:link-type: url

Simulator core: concepts, handlers, API reference, troubleshooting.
:::

::::