Metraipolis › archive › dc_2dfd7621517e

CounterAlign: Counterfactual Supervision for Vision-Language-Action Models — arXiv:2608.21740

Source: arXiv (en)
Original: https://arxiv.org/abs/2608.21740
License (as archived): CC0-1.0
Archived: 2026-10-18T09:01:48.742Z
SHA-256 of the archived text: b0de4d974173ebccf054a62c4210f6dc67b4af34da43465d82c8fa14df585ff2
Stable URL of this page: https://metraipolis.com/documento/dc_2dfd7621517e

How to cite

CounterAlign: Counterfactual Supervision for Vision-Language-Action Models — arXiv:2608.21740 — arXiv (en), https://arxiv.org/abs/2608.21740 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2dfd7621517e (Metraipolis archive dc_2dfd7621517e, sha256 b0de4d974173ebccf054a62c4210f6dc67b4af34da43465d82c8fa14df585ff2)

BibTeX

@misc{dc2dfd7621517e,
  title = {CounterAlign: Counterfactual Supervision for Vision-Language-Action Models — arXiv:2608.21740},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.21740},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2dfd7621517e; sha256 b0de4d974173ebccf054a62c4210f6dc67b4af34da43465d82c8fa14df585ff2},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.21740. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Vision-Language-Action (VLA) models are typically trained with behavior cloning (BC) on expert demonstrations. However, BC provides only positive supervision for expert actions, without explicit negative supervision indicating which actions are instruction-inconsistent or otherwise inappropriate. Reinforcement learning (RL) can provide such corrective signals, but often relies on externally specified rewards or curated non-expert data, both of which are costly to obtain in robotics. We show that offline RL for VLA models need not rely on curated non-expert trajectories: successful expert demonstrations alone can be transformed into dense corrective supervision through instruction relabeling. Specifically, by pairing expert actions with mismatched alternative instructions, we synthesize counterfactual instruction-observation-action tuples from the dataset and combine them with adversarial discriminator training to learn an instruction-grounded reward model for offline RL, without collecting additional rollouts or annotations. On the robustness-focused LIBERO-PRO benchmark, our method improves robustness to object position and task perturbations over a strong state-of-the-art baseline. It also outperforms competitive baselines in real-robot experiments on the TX-G2 (compatible with AGIBot G2). More broadly, our results suggest that, for data-constrained VLA learning, extracting denser supervision from each demonstration can complement collecting additional data.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-08-25 in cs.RO).
Categorie arXiv: Robotics (cs.RO).
Pagina del lavoro: https://arxiv.org/abs/2608.21740 — 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 2026-10-18T08:57:47.841Z dalla pagina https://arxiv.org/catchup/cs.RO/2026-08-25?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).