Metraipolis › archive › dc_07514c0d01fc

Closed-Loop Physics-Based Manipulation Evaluation at a Commodity Price: What 4.4 L4 GPU hours at $3.50 Buys You — arXiv:2611.07441

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.07441
License (as archived)
CC0-1.0
Archived
2026-11-06T14:53:57.332Z
SHA-256 of the archived text
851c017b8e4ab72b149b96c70aac25ce7f61f2eb8580add08b8692dfc9903d20
Stable URL of this page
https://metraipolis.com/documento/dc_07514c0d01fc

How to cite

Closed-Loop Physics-Based Manipulation Evaluation at a Commodity Price: What 4.4 L4 GPU hours at $3.50 Buys You — arXiv:2611.07441 — arXiv (en), https://arxiv.org/abs/2611.07441 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_07514c0d01fc (Metraipolis archive dc_07514c0d01fc, sha256 851c017b8e4ab72b149b96c70aac25ce7f61f2eb8580add08b8692dfc9903d20)
BibTeX
@misc{dc07514c0d01fc,
  title = {Closed-Loop Physics-Based Manipulation Evaluation at a Commodity Price: What 4.4 L4 GPU hours at $3.50 Buys You — arXiv:2611.07441},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.07441},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_07514c0d01fc; sha256 851c017b8e4ab72b149b96c70aac25ce7f61f2eb8580add08b8692dfc9903d20},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.07441. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Running evaluations of contact-rich robotic manipulation at statistical scale has largely been possible only at well-resourced labs, highlighting the need for affordable and efficient options for broader audiences. We present a reproducible evaluation harness wrapper based on a previous study at 200-rollout scale on serverless commodity GPUs, at a cost of approximately $3.50 USD and 38 minutes of wall-clock per replication (4.4 L4-GPU-hours across eight parallel shards). Using the harness, we report a partially controlled negative result on action chunking with sparse data: three independently trained chunked checkpoints produce zero genuine successes across 780 rollouts, while three single-step checkpoints range from 2.6% to 11.8% success rate. A per-step-replanning control partially isolates chunk prediction, rather than open-loop chunk execution as the mechanism, with the failure localized to the training formulation and multi-step prediction. This work further democratizes physics-based policy development and evaluation through our reusable harness.

Tema degli esploratori: statistica e scienza dei dati (elenco arXiv cs.RO, nuovi lavori del giorno).
Categorie arXiv: Robotics (cs.RO).
Pagina del lavoro: https://arxiv.org/abs/2611.07441 — 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 2026-11-06T14:25:57.560Z dalla pagina https://arxiv.org/list/cs.RO/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).