Metraipolis › archive › dc_3bf1088684eb

Rethinking Causal Action Tokenization with Conditional Annealing in Flow Matching — arXiv:2609.35469

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.35469
License (as archived)
CC0-1.0
Archived
2026-10-07T21:03:39.453Z
SHA-256 of the archived text
3459e421c2a84f278c5d19177003783281b794388dbafc5ac59f49a57ae12652
Stable URL of this page
https://metraipolis.com/documento/dc_3bf1088684eb

How to cite

Rethinking Causal Action Tokenization with Conditional Annealing in Flow Matching — arXiv:2609.35469 — arXiv (en), https://arxiv.org/abs/2609.35469 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_3bf1088684eb (Metraipolis archive dc_3bf1088684eb, sha256 3459e421c2a84f278c5d19177003783281b794388dbafc5ac59f49a57ae12652)
BibTeX
@misc{dc3bf1088684eb,
  title = {Rethinking Causal Action Tokenization with Conditional Annealing in Flow Matching — arXiv:2609.35469},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.35469},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_3bf1088684eb; sha256 3459e421c2a84f278c5d19177003783281b794388dbafc5ac59f49a57ae12652},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.35469. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Autoregressive Vision-Language-Action (VLA) models offer a scalable path to robot learning, yet existing action tokenizers treat tokenization as a compression problem, producing representations that are semantically misaligned with the autoregressive backbone. We propose CATok, a causal action tokenizer that reframes tokenization as a causally structured generative process. CATok introduces a conditional annealing mechanism that extracts action tokens by progressively annealing a flow-matching process: each token is conditioned on all preceding tokens and encodes the residual reconstruction signal at a specific noise level, establishing a coarse-to-fine causal token space whose generative semantics are structurally aligned with autoregressive modeling. A token-conditioned flow-matching decoder built on Multimodal Diffusion Transformer (MMDiT) reconstructs continuous action chunks from these discrete tokens with the precision of hybrid diffusion-head architectures. This discrete bottleneck enforces knowledge insulation by design, cleanly separating high-level semantic reasoning from low-level motor execution without requiring explicit attention masking. Extensive evaluations across three simulation benchmarks and real-world robotic manipulation tasks demonstrate that CATok consistently surpasses existing tokenization methods in both reconstruction fidelity-compression tradeoff and inference efficiency, while improving VLA task success rate and training efficiency, establishing a high-performance, scalable foundation for purely autoregressive VLA systems.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.AI).
Categorie arXiv: Robotics (cs.RO) ; Artificial Intelligence (cs.AI); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.35469 — 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 2026-10-07T20:29:03.368Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).