Metraipolis › archive › dc_8bea1ad654cb
BEACON: Belief-Enabled Adaptive CONtrol for Imitation Learning under Uncertainty — arXiv:2609.22730
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.22730
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-09T22:23:35.301Z
- SHA-256 of the archived text
58b0345b44c4772cbc76445205375aa9225979fc866b3cb12a2000295523c4b2- Stable URL of this page
https://metraipolis.com/documento/dc_8bea1ad654cb
How to cite
BEACON: Belief-Enabled Adaptive CONtrol for Imitation Learning under Uncertainty — arXiv:2609.22730 — arXiv (en), https://arxiv.org/abs/2609.22730 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_8bea1ad654cb (Metraipolis archive dc_8bea1ad654cb, sha256 58b0345b44c4772cbc76445205375aa9225979fc866b3cb12a2000295523c4b2)
BibTeX
@misc{dc8bea1ad654cb,
title = {BEACON: Belief-Enabled Adaptive CONtrol for Imitation Learning under Uncertainty — arXiv:2609.22730},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.22730},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_8bea1ad654cb; sha256 58b0345b44c4772cbc76445205375aa9225979fc866b3cb12a2000295523c4b2},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.22730. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Robot manipulation tasks often involve hidden state information that cannot be directly observed and must be inferred through sequential physical interactions. In such partially observable settings, conditioning an imitation learning policy directly on the recent raw observation history leads to poor performance. This is due to state aliasing, wherein identical observations may arise from different hidden states, and the policy receives conflicting action labels for the same input. To enable history-aware disambiguation capability, we propose conditioning a diffusion policy on a structured representation of the hidden states using Bayesian belief that exposes both the current most likely state estimate and the remaining uncertainty. This representation replaces raw history with a structured, compact input, enabling the policy to implicitly modulate between exploratory and exploitative behaviors based on belief uncertainty, without explicit mode switching or reward shaping. We evaluate across two domains with qualitatively different belief representations: a continuous belief for cornstalk gripper alignment via tactile sensing, and a discrete categorical distribution for latched door opening. In both domains, the belief-conditioned policy substantially outperforms the observation-only baseline and approaches privileged ground-truth performance, with ablations illustrating that the policy adapts its exploration behavior depending on the belief uncertainty at inference time. Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-22 in cs.RO). Categorie arXiv: Robotics (cs.RO). Pagina del lavoro: https://arxiv.org/abs/2609.22730 — 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 2026-10-09T21:59:59.466Z dalla pagina https://arxiv.org/catchup/cs.RO/2026-09-22?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).