Metraipolis › archive › dc_a4cfbb5bc075

How To Train Your World Model: Fine-tuning vs RAG for LM-based World Modeling — arXiv:2610.02542

Source: arXiv
Original: https://arxiv.org/abs/2610.02542
License (as archived): CC0-1.0
Archived: 2026-10-05T22:49:02.517Z
SHA-256 of the archived text: 493c08eebf4d58df9171003feb8cc37f6552d7bd7305225989f70dd11df9d2f6
Stable URL of this page: https://metraipolis.com/documento/dc_a4cfbb5bc075

How to cite

How To Train Your World Model: Fine-tuning vs RAG for LM-based World Modeling — arXiv:2610.02542 — arXiv, https://arxiv.org/abs/2610.02542 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_a4cfbb5bc075 (Metraipolis archive dc_a4cfbb5bc075, sha256 493c08eebf4d58df9171003feb8cc37f6552d7bd7305225989f70dd11df9d2f6)

BibTeX

@misc{dca4cfbb5bc075,
  title = {How To Train Your World Model: Fine-tuning vs RAG for LM-based World Modeling — arXiv:2610.02542},
  howpublished = {arXiv},
  url = {https://arxiv.org/abs/2610.02542},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_a4cfbb5bc075; sha256 493c08eebf4d58df9171003feb8cc37f6552d7bd7305225989f70dd11df9d2f6},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.02542. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

World models (WMs) simulate the transition dynamics of environments, enabling agents to plan over the consequences of their actions. In text-based environments, fine-tuning a Language Model (LM) to serve as a WM has emerged as a dominant paradigm. However, despite the widespread success of non-parametric approaches such as Retrieval Augmented Generation (RAG), retrieval for LM-based world modelling remains underexplored. We conduct a systematic evaluation across five diverse environments spanning embodied, web navigation and social settings, comparing fine-tuning and RAG-based approaches for LM-based world modelling. Our study reveals that fine-tuning often outperforms RAG, with fine-tuned WMs enabling agents to obtain higher rewards on 15/20 settings. While both construction paradigms benefit from additional and more diverse exploration, RAG-based approaches prove more data-efficient, and fine-tuning approaches disproportionately benefit from scaling the amount of experience collected. With a focus on RAG-based WMs, we devise a procedure that uses counterfactual intervention to estimate the error rate of the retrieval stage, and show that retrievers consistently surface suboptimal transitions from the experience buffer. Hoping to address this failing, we study a variety of query reformulation strategies, demonstrating that a hierarchical approach outperforms the traditional retrieval pipeline. Finally, we compose our findings into a hybrid world modelling system that parametrically captures core environment dynamics, while learning to rely on retrieval from an actively maintained memory store. Our hybrid system consistently outperforms other methods across multiple environments and models, showcasing the robustness of the approach and the applicability of our findings.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2610.02542 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-05T22:45:55.598Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).