Dreaming in Code for Curriculum Learning in Open-Ended Worlds

TL;DR

This paper introduces DiCode, a framework where foundation models generate environment code to scaffold agent learning in open-ended worlds, leading to improved long-horizon skills and success in complex tasks.

Contribution

The paper presents a novel code-level environment synthesis approach for curriculum learning in open-ended worlds, demonstrating its effectiveness in a new challenging benchmark.

Findings

Achieved 16% improvement in mean return over baselines.

Enabled success on late-game combat tasks where prior methods failed.

Demonstrated long-horizon skill acquisition in complex environments.

Abstract

Open-ended learning frames intelligence as emerging from continual interaction with an ever-expanding space of environments. While recent advances have utilized foundation models to programmatically generate diverse environments, these approaches often focus on discovering isolated behaviors rather than orchestrating sustained progression. In complex open-ended worlds, the large combinatorial space of possible challenges makes it difficult for agents to discover sequences of experiences that remain consistently learnable. To address this, we propose Dreaming in Code (DiCode), a framework in which foundation models synthesize executable environment code to scaffold learning toward increasing competence. In DiCode, "dreaming" takes the form of materializing code-level variations of the world. We instantiate DiCode in Craftax, a challenging open-ended benchmark characterized by rich mechanics and long-horizon progression. Empirically, DiCode enables agents to acquire long-horizon skills, achieving a $16\%$ improvement in mean return over the strongest baseline and non-zero success on late-game combat tasks where prior methods fail. Our results suggest that code-level environment design provides a practical mechanism for curriculum control, enabling the construction of intermediate environments that bridge competence gaps in open-ended worlds. Project page and source code are available at https://konstantinosmitsides.github.io/dreaming-in-code and https://github.com/konstantinosmitsides/dreaming-in-code.