Metraipolis › archive › dc_50a895595bec
Diagnosing and Recovering from Observation-Space Shift at Long-Horizon Skill Seams — arXiv:2610.10810
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.10810
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-09T06:23:52.861Z
- SHA-256 of the archived text
2c21663c5655f06cb3b7b524af60fab334062693cad4142267a19e7b6afe0e37- Stable URL of this page
https://metraipolis.com/documento/dc_50a895595bec
How to cite
Diagnosing and Recovering from Observation-Space Shift at Long-Horizon Skill Seams — arXiv:2610.10810 — arXiv (en), https://arxiv.org/abs/2610.10810 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_50a895595bec (Metraipolis archive dc_50a895595bec, sha256 2c21663c5655f06cb3b7b524af60fab334062693cad4142267a19e7b6afe0e37)
BibTeX
@misc{dc50a895595bec,
title = {Diagnosing and Recovering from Observation-Space Shift at Long-Horizon Skill Seams — arXiv:2610.10810},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.10810},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_50a895595bec; sha256 2c21663c5655f06cb3b7b524af60fab334062693cad4142267a19e7b6afe0e37},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.10810. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Long-horizon robotic manipulation is often built by chaining independently trained skills. Although each skill can be reliable in isolation, performance degrades sharply when skills are chained: each downstream skill must start from the state its predecessor leaves behind rather than from its training distribution. We study this failure mode, Observation-Space Shift (OSS), and ask what causes these skill-seam failures. Using privileged simulator resets, we find that the dominant shift comes from displaced scene state (e.g., an open drawer or secondary objects left behind by earlier skills), not from the robot's joint configuration or the object the downstream skill manipulates. To test this diagnosis, we build a fully learned detect-restore-resume system: a task-progress monitor detects the stall, a learned policy restores the displaced scene components, and seam-robust fine-tuning lets the skill resume. It recovers the seam where every tested alternative fails, which we treat as evidence for the diagnosis rather than as a general-purpose method. On the BOSS-44 benchmark, the system improves full-chain success from 7.6% to 26.5%, a 3.5x improvement over the base policy and 51% of a privileged restoration oracle, whereas best-of-K resampling, a Diffusion Policy, and world-model baselines fail to recover from the evaluated seam states. On a real Franka arm running a fine-tuned $\pi_{0.5}$ policy, the same monitor is limited by exterior-camera observability, yet closing the loop still recovers some otherwise-terminal failures, motivating wrist and gripper sensing. These results suggest that some long-horizon composition failures are better addressed by restoring the scene before resuming the policy than by retrying from an off-support state.
Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Robotics (cs.RO) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.10810 — 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 2026-10-09T05:50:31.896Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).