Game-Guided Skill Discovery through Self-Play for Playable Agent Control
GGSD uses game self-play to learn a small set of human-playable motor skills for embodied agents.
Game-Guided Skill Discovery trains a hierarchical agent through competitive self-play, with a high-level policy choosing from a small discrete skill set and a skill-conditioned low-level policy learning the behaviors. After training, a human can replace the high-level policy and control the agent with those skills. On Ant, Franka-arm, and Unitree G1, people compose the skills to solve unseen tasks such as Maze and CubePush without additional training, and skill transitions produce emergent combo behaviors.
- Hierarchical self-play discovers a compact discrete set of motor skills
- Humans replace the high-level policy and compose the same skills
- Skills solve unseen Maze and CubePush tasks without extra training
- Evaluated on Ant, Franka-arm, and Unitree G1 environments
Full article182 words · extracted from huggingface.co · click to collapse
We present Game-Guided Skill Discovery (GGSD), a framework that uses self-play in games to discover motor skills that are directly playable by humans. Playable skills provide a compact abstraction for controlling embodied agents through a small set of learned behaviors rather than low-level actions. To be effective, these skills should be semantically distinct, interpretable, and expressive; properties that existing unsupervised skill-discovery methods often fail to achieve simultaneously. GGSD achieves these desiderata by grounding skill discovery in competitive gameplay. A hierarchical agent competes against its past selves, with a high-level policy selecting from a small discrete skill set and a skill-conditioned low-level policy learning the corresponding behaviors. After training, a human can replace the high-level policy and directly control the agent through the same discrete skills. Despite the small number of high-level actions, skill transitions give rise to emergent combo behaviors, expanding expressivity beyond individual primitives. Across Ant, Franka-arm, and Unitree G1 environments, we show that GGSD produces human-playable skills that humans can compose to solve unseen tasks, such as Maze and CubePush, without additional training. An interactive demo is available at https://ggsd-demo.github.io.
Text extracted automatically; images, tables and formatting may be missing. Original: https://huggingface.co/papers/2609.40137