21 lines (13 loc) · 1004 Bytes

Acknowledgements / Notice

This repository is a release-oriented, reading-friendly packaging of the code for:

Imagine-then-Plan: Agent Learning from Adaptive Lookahead with World Models (arXiv:2601.08955)
- Paper: https://arxiv.org/abs/2601.08955
- PDF: https://arxiv.org/pdf/2601.08955

We gratefully acknowledge the open-source projects and environments that inspired or supported this work:

ETO (repository organization style reference): we follow a similar “few top-level modules + clear boundaries” layout.
Hugging Face Transformers / Accelerate / PEFT / TRL for model loading, fine-tuning, and training utilities.
ALFWorld and TextWorld for the household text environment benchmark.
ScienceWorld for the science-based text environment benchmark.

Licenses:

This repository is released under the MIT License (see LICENSE).
Third-party projects listed above have their own licenses; please refer to the original upstream repositories.