Metraipolis › archive › dc_4826323c5bbd
Changepoint-Aware World Models: Detecting Dynamics Shifts and Recovering by Forgetting Stale Replay in Model-Based RL — arXiv:2609.18950
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.18950
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-15T15:04:56.719Z
- SHA-256 of the archived text
a7b3a78b3fc4418dbe22368340babb4634965cf03082d2f9fb351d3c0ef7e98d- Stable URL of this page
https://metraipolis.com/documento/dc_4826323c5bbd
How to cite
Changepoint-Aware World Models: Detecting Dynamics Shifts and Recovering by Forgetting Stale Replay in Model-Based RL — arXiv:2609.18950 — arXiv (en), https://arxiv.org/abs/2609.18950 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_4826323c5bbd (Metraipolis archive dc_4826323c5bbd, sha256 a7b3a78b3fc4418dbe22368340babb4634965cf03082d2f9fb351d3c0ef7e98d)
BibTeX
@misc{dc4826323c5bbd,
title = {Changepoint-Aware World Models: Detecting Dynamics Shifts and Recovering by Forgetting Stale Replay in Model-Based RL — arXiv:2609.18950},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.18950},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_4826323c5bbd; sha256 a7b3a78b3fc4418dbe22368340babb4634965cf03082d2f9fb351d3c0ef7e98d},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.18950. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
A robot's learned model of its own dynamics is only valid until those dynamics change: actuators wear, payloads shift, and joints stiffen. A model-based agent that keeps training as if nothing happened adapts slowly, dragged back by a replay buffer full of stale experience. We present Changepoint-Aware World Models (CAWM), a DreamerV3 agent that detects an abrupt dynamics shift from its own internal prediction error, using an online CUSUM test against a rolling baseline that fires only on abrupt change rather than on slow learning drift. It then forgets stale replay, keeping the learned representation while flushing obsolete data. On simulated locomotion under two robot-relevant shifts, doubled gravity and halved actuator gain, CAWM recovers substantially faster than passive retraining. It also beats a strong baseline that respawns a fresh dynamics model on detection, the deep-world-model analogue of model-bank methods. With the response triggered at the shift, CAWM gains +95 to +153 return in the first 30k post-shift frames over three seeds, while matching that respawn at asymptote. Running the detector in closed loop reproduces this gain on the gravity shift. The benefit holds across both shift types, and is largest when the shift is severe enough that old data is genuinely obsolete. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-17 in cs.LG). Categorie arXiv: Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2609.18950 — 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 2026-10-15T14:08:19.159Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-17?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).