Metraipolis › archive › dc_ee9f07245654

Tactile Curiosity Drives Robot Interaction — arXiv:2609.40134

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.40134
License (as archived)
CC0-1.0
Archived
2026-10-07T12:46:11.627Z
SHA-256 of the archived text
85c89d84b1fc01f4793259e6ff782b40434d7c2641c2676a8df32054ae9183a9
Stable URL of this page
https://metraipolis.com/documento/dc_ee9f07245654

How to cite

Tactile Curiosity Drives Robot Interaction — arXiv:2609.40134 — arXiv (en), https://arxiv.org/abs/2609.40134 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ee9f07245654 (Metraipolis archive dc_ee9f07245654, sha256 85c89d84b1fc01f4793259e6ff782b40434d7c2641c2676a8df32054ae9183a9)
BibTeX
@misc{dcee9f07245654,
  title = {Tactile Curiosity Drives Robot Interaction — arXiv:2609.40134},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.40134},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ee9f07245654; sha256 85c89d84b1fc01f4793259e6ff782b40434d7c2641c2676a8df32054ae9183a9},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.40134. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Mastering robot manipulation skills via reinforcement learning (RL) remains largely sample-inefficient. The most common RL algorithms rely on random action sampling to discover new strategies, resulting in agents that allocate most of their training budget to motions in free space, away from the contacts from which manipulation skills emerge. Existing intrinsic motivation methods based on model disagreement or epistemic uncertainty improve on isotropic noise, but they can also reward uncertainty in functionally irrelevant transitions, such as erratic motions in free space. In this work, we argue that tactile feedback provides a natural signal for exploration, and introduce TacEx, a framework that incorporates touch into epistemic uncertainty-driven exploration by decomposing model uncertainty across sensory modalities and directing curiosity toward the tactile channel. By anchoring curiosity to the sense of touch, TacEx drives the robot to discover complex contact dynamics, learning to manipulate and grasp objects without task rewards or expert demonstrations during exploration. The interaction-dense dataset collected through this tactile-driven curiosity supports offline learning of downstream pick-and-place policies without additional environment interaction. We further use tactile-driven exploration to post-train vision-language-action (VLA) models. Although the VLAs are initially pre-trained without tactile feedback, post-training with TacEx substantially improves downstream performance while remaining highly sample-efficient.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-10-01 in cs.AI).
Categorie arXiv: Robotics (cs.RO) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.40134 — 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 2026-10-07T12:14:14.223Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-10-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).