Metraipolis › archive › dc_0be797281f89

WorldCycle: Self-Verifiable Reinforcement Learning for Long-Horizon Video World Models — arXiv:2608.04964

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.04964
License (as archived)
CC0-1.0
Archived
2026-10-31T11:13:13.747Z
SHA-256 of the archived text
1981a5e1bee45f5b93f3794f577df8e0158116adac6ef6b60e46d02d2e51c93a
Stable URL of this page
https://metraipolis.com/documento/dc_0be797281f89

How to cite

WorldCycle: Self-Verifiable Reinforcement Learning for Long-Horizon Video World Models — arXiv:2608.04964 — arXiv (en), https://arxiv.org/abs/2608.04964 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_0be797281f89 (Metraipolis archive dc_0be797281f89, sha256 1981a5e1bee45f5b93f3794f577df8e0158116adac6ef6b60e46d02d2e51c93a)
BibTeX
@misc{dc0be797281f89,
  title = {WorldCycle: Self-Verifiable Reinforcement Learning for Long-Horizon Video World Models — arXiv:2608.04964},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.04964},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_0be797281f89; sha256 1981a5e1bee45f5b93f3794f577df8e0158116adac6ef6b60e46d02d2e51c93a},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.04964. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Interactive video world models are essential for long-horizon planning and exploration, yet they suffer from compounding errors. Post-training methods such as reinforcement learning (RL) can improve these models, but they hit a verification bottleneck: for arbitrary action sequences, no ground-truth future state exists to measure long-term drift. Our key insight is that reversible action cycles make this verification possible: a sequence composed with its inverse must analytically return to the initial state, yielding annotation-free supervision on long-horizon correctness. Building on this, we introduce WorldCycle, a self-verifiable RL framework that constructs closed action cycles and their repeated executions from ordinary action sequences, and optimizes two complementary rewards: a spatial closure reward enforcing symmetry between mirrored forward and reverse segments, and a temporal consistency reward aligning states across repeated cycle executions. These rewards force the model to learn actions as consistent state operators rather than memorized temporal patterns, and extend naturally to out-of-distribution composite action cycles that the base model handles poorly. We further release CycleBench, a diagnostic benchmark for state-returning ability under complex action structures. WorldCycle reduces state returning drift by up to 44% and lifts composite-action accuracy nearly 4x over the base model, providing a vital foundation for physically grounded world models.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-06 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2608.04964 — 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 2026-10-31T11:02:13.434Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-06?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).