---
title: README
emoji: 📚
colorFrom: green
colorTo: pink
sdk: static
pinned: false
---

# AceAction Research

**Learning to Act in Interactive Worlds.**

AceAction is a research initiative exploring how intelligent agents understand goals, perceive dynamic environments, and translate intentions into actions. Our interests span **vision-language-action (VLA) learning**, **action-conditioned world modeling**, and **long-horizon decision-making** in interactive 3D worlds.

## 📄 Papers

### [AceAction-5M](https://arxiv.org/abs/2611.04241) · arXiv 2026

**Hierarchical Goals for Generalist Action Learning Across 3D Game Worlds**

A multi-game action-learning dataset with hierarchically structured goals, subgoals, and skills across **42 games** and over **5 million annotated episodes**.



### [3A-VLA](https://arxiv.org/abs/2602.02212) · NeurIPS 2026

**Abstraction-Aligned Action Learning for Vision-Language Agents in 3D Game Worlds**