Metraipolis › archive › dc_3ffe27b33964

Direction-Scale Decomposition in Action Representation: Rethinking What to Tokenize for Vision-Language-Action Models — arXiv:2609.28865

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.28865
License (as archived)
CC0-1.0
Archived
2026-10-08T06:44:16.527Z
SHA-256 of the archived text
ca83161b86b80351c9100a9b3f3f3d7ae6c9c59195c076f42dcaf930fd5f7080
Stable URL of this page
https://metraipolis.com/documento/dc_3ffe27b33964

How to cite

Direction-Scale Decomposition in Action Representation: Rethinking What to Tokenize for Vision-Language-Action Models — arXiv:2609.28865 — arXiv (en), https://arxiv.org/abs/2609.28865 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_3ffe27b33964 (Metraipolis archive dc_3ffe27b33964, sha256 ca83161b86b80351c9100a9b3f3f3d7ae6c9c59195c076f42dcaf930fd5f7080)
BibTeX
@misc{dc3ffe27b33964,
  title = {Direction-Scale Decomposition in Action Representation: Rethinking What to Tokenize for Vision-Language-Action Models — arXiv:2609.28865},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.28865},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_3ffe27b33964; sha256 ca83161b86b80351c9100a9b3f3f3d7ae6c9c59195c076f42dcaf930fd5f7080},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.28865. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Action representation plays a central role in discrete-token vision-language-action (VLA) learning but remains underexamined. Under conventional pose-increment representations, action tokens are sensitive to execution speed and dataset-specific normalization, potentially obscuring geometric structure shared across demonstrations and datasets. We introduce Direction-Scale Decomposition (DSD), an action representation that decomposes translation and rotation increments into direction and scale components before tokenization. DSD isolates motion direction while retaining magnitudes in separate scale channels. We evaluate DSD with uniform binning (BIN) and BEAST, a B-spline-based tokenizer, in simulation and real-world manipulation under both single-dataset and mixed-dataset training. On LIBERO, DSD improves average success rates with both tokenizers. On SimplerEnv, DSD-BIN outperforms BIN by 10.3 percentage points in overall success rate under mixed-dataset training. Real-robot experiments further show gains both with and without robotics pretraining. These results support DSD as an effective action representation for discrete-token VLA models and suggest its potential to mitigate performance degradation when training on large and diverse dataset mixtures. Our project page with additional resources is available at this https URL

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-25 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Robotics (cs.RO).
Pagina del lavoro: https://arxiv.org/abs/2609.28865 — 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 2026-10-08T06:09:29.788Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-09-25?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).