Metraipolis › archive › dc_985a127b618c

Reflex: Fast Robot Policy Serving by Reusing and Correcting Past Actions — arXiv:2611.00524

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.00524
License (as archived)
CC0-1.0
Archived
2026-11-03T10:55:36.358Z
SHA-256 of the archived text
9ca42f00b68c199218509455da1729b82b15fac3a7d5fd9e5aec0780e46cf768
Stable URL of this page
https://metraipolis.com/documento/dc_985a127b618c

How to cite

Reflex: Fast Robot Policy Serving by Reusing and Correcting Past Actions — arXiv:2611.00524 — arXiv (en), https://arxiv.org/abs/2611.00524 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_985a127b618c (Metraipolis archive dc_985a127b618c, sha256 9ca42f00b68c199218509455da1729b82b15fac3a7d5fd9e5aec0780e46cf768)
BibTeX
@misc{dc985a127b618c,
  title = {Reflex: Fast Robot Policy Serving by Reusing and Correcting Past Actions — arXiv:2611.00524},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.00524},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_985a127b618c; sha256 9ca42f00b68c199218509455da1729b82b15fac3a7d5fd9e5aec0780e46cf768},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.00524. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Modern robot policies often perform expensive inference to generate action chunks, making serving a latency bottleneck for real-time control and a throughput bottleneck for multi-robot serving. Building on the success of caching in image, video, and audio generation, recent robot-policy advances have reported promising speedups by reusing intermediate denoising states to avoid redundant computation. Surprisingly, we find that such reuse often offers little advantage over simply running fewer denoising steps from fresh noise, without any cache. Moreover, accelerating denoising leaves the expensive policy backbone intact, which can account for 70% of GPU time. We present Reflex, a serving layer that turns past policy executions into a backbone-free path for future decisions. Rather than reusing intermediate computations, Reflex caches complete action chunks to bypass both the backbone and action head. This introduces two challenges: similar observations can demand different actions, and small action deviations (errors) in reuse can compound into execution stalls. Reflex addresses these through two-level cache correction. At the action level, it learns action-aware similarity from past policy decisions, assembles compatible chunks, and applies lightweight residual correction to limit action mismatch. At the execution level, it captures task progress and adaptively refreshes the original policy to limit accumulated errors and balance inference cost. Without modifying the underlying policy, Reflex improves inference latency by 1.4x to 11x and throughput by 2.0x to 2.2x on representative workloads and real robots, while preserving comparable task success.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Robotics (cs.RO) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2611.00524 — 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 2026-11-03T10:49:38.489Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).